Compare commits
11
Commits
| Author | SHA1 | Date | |
|---|---|---|---|
|
|
f541ec4260 | ||
|
|
f549f020e1 | ||
|
|
0ff80a0fd3 | ||
|
|
4df73367fb | ||
|
|
a437fa87d3 | ||
|
|
32a1583efd | ||
|
|
03676432bd | ||
|
|
8d6b9212c1 | ||
|
|
18f172fe96 | ||
|
|
b576a9cdf3 | ||
|
|
5973920894 |
@@ -22,18 +22,18 @@ This document provides essential context for AI agents working with URLNotesGrab
|
||||
```powershell
|
||||
dotnet build
|
||||
dotnet run # Process all files in input directory
|
||||
dotnet run -- -parse [blogname] # Process specific blog
|
||||
dotnet run -- -test [blogname] [postID] # Test API for specific post
|
||||
dotnet run -- --parse [blogname] # Process specific blog
|
||||
dotnet run -- --test [blogname] [postID] # Test API for specific post
|
||||
```
|
||||
|
||||
### Command-Line Interface
|
||||
- `-parse [blogname]`: Parse text files for specific blog
|
||||
- `-test [blogname] [postID]`: Test API note collection
|
||||
- `-posts`: Export post blogs to file
|
||||
- `-blogs`: Export blog list to file
|
||||
- `-collect`: Collect notes for all posts in DB
|
||||
- `-blogsR`: Export reply blogs to file
|
||||
- `-blogsO [start] [stop]`: Export blogs within range
|
||||
- `--parse [blogname]`: Parse text files for specific blog
|
||||
- `--test [blogname] [postID]`: Test API note collection
|
||||
- `--posts`: Export post blogs to file
|
||||
- `--blogs`: Export blog list to file
|
||||
- `--collect`: Collect notes for all posts in DB
|
||||
- `--blogsR`: Export reply blogs to file
|
||||
- `--blogsO [start] [stop]`: Export blogs within range
|
||||
|
||||
## Project Conventions
|
||||
|
||||
|
||||
BIN
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.
+108
-146
@@ -139,7 +139,7 @@ namespace URLNotesGrabberCORE
|
||||
if (_savedImportPragmas != null)
|
||||
return;
|
||||
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
try
|
||||
{
|
||||
connection.Open();
|
||||
@@ -175,10 +175,6 @@ namespace URLNotesGrabberCORE
|
||||
{
|
||||
Console.WriteLine($"[SQLite Import Mode] Failed to enable import pragmas: {ex.Message}");
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
@@ -190,7 +186,7 @@ namespace URLNotesGrabberCORE
|
||||
if (_savedImportPragmas == null)
|
||||
return;
|
||||
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
try
|
||||
{
|
||||
connection.Open();
|
||||
@@ -217,7 +213,6 @@ namespace URLNotesGrabberCORE
|
||||
finally
|
||||
{
|
||||
_savedImportPragmas = null;
|
||||
connection.Close();
|
||||
}
|
||||
}
|
||||
}
|
||||
@@ -252,7 +247,7 @@ namespace URLNotesGrabberCORE
|
||||
public static void EnsureReplyTextColumnExists(string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
|
||||
try
|
||||
{
|
||||
@@ -296,16 +291,12 @@ namespace URLNotesGrabberCORE
|
||||
// Breakpoint here
|
||||
Console.WriteLine($"Error checking/creating replyText column: {ex.Message}");
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
}
|
||||
|
||||
public static void EnsureBlogsLikesColumnsExist(string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
|
||||
try
|
||||
{
|
||||
@@ -391,10 +382,6 @@ namespace URLNotesGrabberCORE
|
||||
{
|
||||
Console.WriteLine($"Error mapping Blogs likes columns: {ex.Message}");
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
}
|
||||
|
||||
#region Adds
|
||||
@@ -518,8 +505,12 @@ namespace URLNotesGrabberCORE
|
||||
{
|
||||
if (ownsConnection) connection.Open();
|
||||
|
||||
string updateSql = "UPDATE Posts SET hasImage = " + (hasImage ? 1 : 0) + ", DateModified = '" + DateTime.Now.ToString("yyyy-MM-dd HH:mm:ss") + "' WHERE blogName = '" + blogName + "' AND postID = '" + postID + "'";
|
||||
SQLiteCommand updateCommand = new SQLiteCommand(updateSql, connection);
|
||||
string updateSql = "UPDATE Posts SET hasImage = @hasImage, DateModified = @DateModified WHERE blogName = @blogName AND postID = @postID";
|
||||
using SQLiteCommand updateCommand = new SQLiteCommand(updateSql, connection);
|
||||
updateCommand.Parameters.AddWithValue("@hasImage", hasImage ? 1 : 0);
|
||||
updateCommand.Parameters.AddWithValue("@DateModified", DateTime.Now.ToString("yyyy-MM-dd HH:mm:ss"));
|
||||
updateCommand.Parameters.AddWithValue("@blogName", blogName);
|
||||
updateCommand.Parameters.AddWithValue("@postID", postID);
|
||||
|
||||
updateCommand.ExecuteNonQuery();
|
||||
}
|
||||
@@ -565,7 +556,7 @@ namespace URLNotesGrabberCORE
|
||||
public static void AddAPICount(string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
|
||||
try
|
||||
{
|
||||
@@ -586,10 +577,6 @@ namespace URLNotesGrabberCORE
|
||||
// Breakpoint here
|
||||
//Console.WriteLine(ex.Message);
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
}
|
||||
|
||||
public static bool AddNote(string rootBlogName, string noteBlogName, long postID, long timestamp, string type, string? DBPath = null)
|
||||
@@ -598,7 +585,7 @@ namespace URLNotesGrabberCORE
|
||||
//try { AddPost(rootBlogName, postID, DBPath); } catch { }
|
||||
try { AddBlog(noteBlogName, false, DBPath); } catch { }
|
||||
|
||||
SQLiteConnection connection2 = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection2 = new SQLiteConnection("Data Source=" + DBPath);
|
||||
|
||||
try
|
||||
{
|
||||
@@ -657,10 +644,6 @@ namespace URLNotesGrabberCORE
|
||||
Console.WriteLine("^^^^^ - SHORTCUT");
|
||||
}
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection2.Close();
|
||||
}
|
||||
return false;
|
||||
}
|
||||
#endregion Adds
|
||||
@@ -676,7 +659,7 @@ namespace URLNotesGrabberCORE
|
||||
public static List<Tuple<string, long, long, long>> GetPosts(bool withoutNotesOnly = false, DateTime? beforeDate = null, string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
List<Tuple<string, long, long, long>> posts = new List<Tuple<string, long, long, long>>();
|
||||
|
||||
try
|
||||
@@ -854,17 +837,13 @@ namespace URLNotesGrabberCORE
|
||||
if (ex.Message != "constraint failed\r\nUNIQUE constraint failed: Posts.BlogName, Posts.PostID")
|
||||
Console.WriteLine(ex.Message);
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
return posts;
|
||||
}
|
||||
|
||||
public static List<Tuple<string, long>> GetReplies(string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
List<Tuple<string, long>> posts = new List<Tuple<string, long>>();
|
||||
|
||||
try
|
||||
@@ -898,17 +877,13 @@ namespace URLNotesGrabberCORE
|
||||
if (ex.Message != "constraint failed\r\nUNIQUE constraint failed: Posts.BlogName, Posts.PostID")
|
||||
Console.WriteLine(ex.Message);
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
return posts;
|
||||
}
|
||||
|
||||
public static List<Tuple<string, long>> GetRepliesWithMissingText(string? DBPath = null, int limit = 50)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
List<Tuple<string, long>> posts = new List<Tuple<string, long>>();
|
||||
|
||||
try
|
||||
@@ -949,17 +924,13 @@ namespace URLNotesGrabberCORE
|
||||
// Breakpoint here
|
||||
Console.WriteLine($"Error getting replies with missing text: {ex.Message}");
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
return posts;
|
||||
}
|
||||
|
||||
public static List<Tuple<string, long, long>> GetRepliesWithFilledText(string? DBPath = null, int? limit = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
List<Tuple<string, long, long>> posts = new List<Tuple<string, long, long>>();
|
||||
|
||||
try
|
||||
@@ -1011,17 +982,13 @@ namespace URLNotesGrabberCORE
|
||||
// Breakpoint here
|
||||
Console.WriteLine($"Error getting replies with filled text: {ex.Message}");
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
return posts;
|
||||
}
|
||||
|
||||
public static int GetAPICount(string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
int count = 0;
|
||||
|
||||
try { AddAPICount(); } catch { }
|
||||
@@ -1050,17 +1017,13 @@ namespace URLNotesGrabberCORE
|
||||
if (ex.Message != "constraint failed\r\nUNIQUE constraint failed: Posts.BlogName, Posts.PostID")
|
||||
Console.WriteLine(ex.Message);
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
return count;
|
||||
}
|
||||
|
||||
public static List<Tuple<string, int, long, long>> GetBlogsForLikes(string specificBlog = null, int cooldownDays = 7, bool ignoreCooldown = false, string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
List<Tuple<string, int, long, long>> blogs = new List<Tuple<string, int, long, long>>();
|
||||
|
||||
try
|
||||
@@ -1136,17 +1099,13 @@ namespace URLNotesGrabberCORE
|
||||
{
|
||||
Console.WriteLine($"Error fetching blogs for likes: {ex.Message}");
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
return blogs;
|
||||
}
|
||||
|
||||
public static List<string> GetBlogs(bool reblogsOnly, int from, int to, int top, string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
List<string> blogs = new List<string>();
|
||||
|
||||
try
|
||||
@@ -1181,17 +1140,13 @@ namespace URLNotesGrabberCORE
|
||||
if (ex.Message != "constraint failed\r\nUNIQUE constraint failed: Posts.BlogName, Posts.PostID")
|
||||
Console.WriteLine(ex.Message);
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
return blogs;
|
||||
}
|
||||
|
||||
public static List<string> GetBlogsAll(bool reblogsOnly, int from, int to, int top, string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
List<string> blogs = new List<string>();
|
||||
|
||||
try
|
||||
@@ -1226,51 +1181,39 @@ namespace URLNotesGrabberCORE
|
||||
if (ex.Message != "constraint failed\r\nUNIQUE constraint failed: Posts.BlogName, Posts.PostID")
|
||||
Console.WriteLine(ex.Message);
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
return blogs;
|
||||
}
|
||||
public static IEnumerable<List<string>> GetAllPostTextColumns(string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
connection.Open();
|
||||
try
|
||||
string sql = "SELECT BlogName, reblogURL, PostURL, Slug, ReblogKey, ReblogName, Summary, Quote, Body, Tags, Link, PhotoURL, PhotoCaption, DownloadedFiles, AudioCaption, Question, Answer, Title, RootBlogName, RootURL FROM Posts WHERE IFNULL(DownloadedFiles, '.') = '.'";
|
||||
using (SQLiteCommand command = new SQLiteCommand(sql, connection))
|
||||
{
|
||||
string sql = "SELECT BlogName, reblogURL, PostURL, Slug, ReblogKey, ReblogName, Summary, Quote, Body, Tags, Link, PhotoURL, PhotoCaption, DownloadedFiles, AudioCaption, Question, Answer, Title, RootBlogName, RootURL FROM Posts WHERE IFNULL(DownloadedFiles, '.') = '.'";
|
||||
using (SQLiteCommand command = new SQLiteCommand(sql, connection))
|
||||
using (SQLiteDataReader reader = command.ExecuteReader())
|
||||
{
|
||||
using (SQLiteDataReader reader = command.ExecuteReader())
|
||||
while (reader.Read())
|
||||
{
|
||||
while (reader.Read())
|
||||
List<string> rowTexts = new List<string>();
|
||||
for (int i = 0; i < reader.FieldCount; i++)
|
||||
{
|
||||
List<string> rowTexts = new List<string>();
|
||||
for (int i = 0; i < reader.FieldCount; i++)
|
||||
if (!reader.IsDBNull(i))
|
||||
{
|
||||
if (!reader.IsDBNull(i))
|
||||
var val = reader.GetValue(i);
|
||||
if (val is string str && !string.IsNullOrWhiteSpace(str) && str != ".")
|
||||
{
|
||||
var val = reader.GetValue(i);
|
||||
if (val is string str && !string.IsNullOrWhiteSpace(str) && str != ".")
|
||||
{
|
||||
rowTexts.Add(str);
|
||||
}
|
||||
rowTexts.Add(str);
|
||||
}
|
||||
}
|
||||
if (rowTexts.Count > 0)
|
||||
{
|
||||
yield return rowTexts;
|
||||
}
|
||||
}
|
||||
if (rowTexts.Count > 0)
|
||||
{
|
||||
yield return rowTexts;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
connection.Dispose();
|
||||
}
|
||||
}
|
||||
#endregion Gets
|
||||
|
||||
@@ -1280,7 +1223,7 @@ namespace URLNotesGrabberCORE
|
||||
public static void UpdatePostMarkNotesCollected(string blogName, long postID, string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
|
||||
try
|
||||
{
|
||||
@@ -1303,16 +1246,12 @@ namespace URLNotesGrabberCORE
|
||||
if (ex.Message != "constraint failed\r\nUNIQUE constraint failed: Posts.BlogName, Posts.PostID")
|
||||
Console.WriteLine(ex.Message);
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
}
|
||||
|
||||
public static void UpdatePostMarkNotFound(string blogName, long postID, string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
|
||||
try
|
||||
{
|
||||
@@ -1336,10 +1275,69 @@ namespace URLNotesGrabberCORE
|
||||
if (ex.Message != "constraint failed\r\nUNIQUE constraint failed: Posts.BlogName, Posts.PostID")
|
||||
Console.WriteLine(ex.Message);
|
||||
}
|
||||
finally
|
||||
}
|
||||
|
||||
// ----- CollectRunState: tracks the frozen cutoff + completion flag for a managed "-collect 0" full re-check run -----
|
||||
// Single-row table (Id = 1), mirroring the ApiKeyPoolMeta pattern. Lets an interrupted run resume against the
|
||||
// same cutoff and lets a completed run stop instead of restarting on the next launch.
|
||||
|
||||
public static void EnsureCollectRunStateTableExists(string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
connection.Open();
|
||||
string sql = "CREATE TABLE IF NOT EXISTS CollectRunState (" +
|
||||
"Id INTEGER PRIMARY KEY CHECK (Id = 1), " +
|
||||
"RunCutoff INTEGER DEFAULT 0, " +
|
||||
"RunComplete INTEGER DEFAULT 1, " +
|
||||
"RunStarted TEXT, " +
|
||||
"RunCompletedAt TEXT)";
|
||||
using SQLiteCommand command = new SQLiteCommand(sql, connection);
|
||||
command.ExecuteNonQuery();
|
||||
}
|
||||
|
||||
// Returns (RunCutoff unix seconds, RunComplete) for the single run-state row, or null if no row exists yet.
|
||||
public static (long cutoff, bool complete)? GetCollectRunState(string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
connection.Open();
|
||||
using SQLiteCommand command = new SQLiteCommand("SELECT RunCutoff, RunComplete FROM CollectRunState WHERE Id = 1", connection);
|
||||
using SQLiteDataReader reader = command.ExecuteReader();
|
||||
if (reader.Read())
|
||||
{
|
||||
connection.Close();
|
||||
long cutoff = Convert.ToInt64(reader.GetValue(0));
|
||||
bool complete = Convert.ToInt64(reader.GetValue(1)) != 0;
|
||||
return (cutoff, complete);
|
||||
}
|
||||
return null;
|
||||
}
|
||||
|
||||
// Start (or restart) a managed run: freeze the cutoff and mark the run in progress.
|
||||
public static void BeginCollectRun(long cutoffUnixSeconds, string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
connection.Open();
|
||||
string sql = "INSERT INTO CollectRunState (Id, RunCutoff, RunComplete, RunStarted, RunCompletedAt) " +
|
||||
"VALUES (1, @cutoff, 0, @started, NULL) " +
|
||||
"ON CONFLICT(Id) DO UPDATE SET RunCutoff = @cutoff, RunComplete = 0, RunStarted = @started, RunCompletedAt = NULL";
|
||||
using SQLiteCommand command = new SQLiteCommand(sql, connection);
|
||||
command.Parameters.AddWithValue("@cutoff", cutoffUnixSeconds);
|
||||
command.Parameters.AddWithValue("@started", DateTime.Now.ToString("yyyy-MM-dd HH:mm:ss"));
|
||||
command.ExecuteNonQuery();
|
||||
}
|
||||
|
||||
// Mark the active managed run complete so the next launch starts fresh instead of resuming.
|
||||
public static void CompleteCollectRun(string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
connection.Open();
|
||||
string sql = "UPDATE CollectRunState SET RunComplete = 1, RunCompletedAt = @completedAt WHERE Id = 1";
|
||||
using SQLiteCommand command = new SQLiteCommand(sql, connection);
|
||||
command.Parameters.AddWithValue("@completedAt", DateTime.Now.ToString("yyyy-MM-dd HH:mm:ss"));
|
||||
command.ExecuteNonQuery();
|
||||
}
|
||||
|
||||
public static void UpdatePostSetDate(string blogName, long postID, string postDate, string? DBPath = null)
|
||||
@@ -1388,7 +1386,7 @@ namespace URLNotesGrabberCORE
|
||||
|
||||
Console.WriteLine("{2}\t{0}\t{1}", UnixTimeStampToDateTime(timestamp), noteBlogName, type);
|
||||
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
|
||||
try
|
||||
{
|
||||
@@ -1415,10 +1413,6 @@ namespace URLNotesGrabberCORE
|
||||
return true;
|
||||
}
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
return false;
|
||||
}
|
||||
|
||||
@@ -1534,7 +1528,7 @@ namespace URLNotesGrabberCORE
|
||||
public static void UpdateBlogOutput(string blogName, string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
|
||||
try
|
||||
{
|
||||
@@ -1554,16 +1548,12 @@ namespace URLNotesGrabberCORE
|
||||
if (ex.Message != "constraint failed\r\nUNIQUE constraint failed: Posts.BlogName, Posts.PostID")
|
||||
Console.WriteLine(ex.Message);
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
}
|
||||
|
||||
public static void UpdateBlogLikesStatus(string blogName, int likesPulled, long likesCursor, string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
try
|
||||
{
|
||||
connection.Open();
|
||||
@@ -1581,10 +1571,6 @@ namespace URLNotesGrabberCORE
|
||||
{
|
||||
Console.WriteLine($"Error updating blog likes status: {ex.Message}");
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
}
|
||||
|
||||
// Bumps the high-water mark for a blog. Used during Branch A (initial backfill) when we
|
||||
@@ -1594,7 +1580,7 @@ namespace URLNotesGrabberCORE
|
||||
{
|
||||
if (newestTimestamp <= 0) return;
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
try
|
||||
{
|
||||
connection.Open();
|
||||
@@ -1614,10 +1600,6 @@ namespace URLNotesGrabberCORE
|
||||
{
|
||||
Console.WriteLine($"Error updating blog likes newest timestamp: {ex.Message}");
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
}
|
||||
|
||||
// Called at the end of a refresh pass (Branch B). Bumps the high-water mark, stamps the
|
||||
@@ -1625,7 +1607,7 @@ namespace URLNotesGrabberCORE
|
||||
public static void UpdateBlogLikesRefreshStatus(string blogName, long newestTimestamp, int newCount, string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
try
|
||||
{
|
||||
connection.Open();
|
||||
@@ -1648,16 +1630,12 @@ namespace URLNotesGrabberCORE
|
||||
{
|
||||
Console.WriteLine($"Error updating blog likes refresh status: {ex.Message}");
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
}
|
||||
|
||||
public static int UpdateAPICount(string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
|
||||
int APICount = DataAccess.GetAPICount();
|
||||
APICount++;
|
||||
@@ -1680,10 +1658,6 @@ namespace URLNotesGrabberCORE
|
||||
if (ex.Message != "constraint failed\r\nUNIQUE constraint failed: Posts.BlogName, Posts.PostID")
|
||||
Console.WriteLine(ex.Message);
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
|
||||
return APICount;
|
||||
}
|
||||
@@ -1691,7 +1665,7 @@ namespace URLNotesGrabberCORE
|
||||
public static int UpdateNoteReplyText(string rootBlogName, long postID, string noteBlogName, long timestamp, string replyText, string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
int rowsAffected = 0;
|
||||
|
||||
// A bare "." collides with the "needs processing" sentinel in GetRepliesWithFilledText, which would loop the post forever. Store as ". " so the data is preserved but no longer matches the sentinel.
|
||||
@@ -1734,17 +1708,13 @@ namespace URLNotesGrabberCORE
|
||||
Console.WriteLine($"[UpdateNoteReplyText] Error updating reply text: {ex.Message}");
|
||||
Console.WriteLine($"[UpdateNoteReplyText] StackTrace: {ex.StackTrace}");
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
return rowsAffected;
|
||||
}
|
||||
|
||||
public static int UpdateAllNoteReplyTextForPost(string rootBlogName, long postID, string replyText, string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
|
||||
try
|
||||
{
|
||||
@@ -1780,10 +1750,6 @@ namespace URLNotesGrabberCORE
|
||||
Console.WriteLine($"[UpdateAllNoteReplyTextForPost] StackTrace: {ex.StackTrace}");
|
||||
return 0;
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
}
|
||||
#endregion Updates
|
||||
|
||||
@@ -1794,7 +1760,7 @@ namespace URLNotesGrabberCORE
|
||||
public static void EnsureTTFileHelperColumnsExist(string? DBPath = null)
|
||||
{
|
||||
DBPath ??= GetDefaultDbPath();
|
||||
SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
using SQLiteConnection connection = new SQLiteConnection("Data Source=" + DBPath);
|
||||
|
||||
try
|
||||
{
|
||||
@@ -1840,10 +1806,6 @@ namespace URLNotesGrabberCORE
|
||||
{
|
||||
Console.WriteLine($"Error ensuring TTFileHelper columns: {ex.Message}");
|
||||
}
|
||||
finally
|
||||
{
|
||||
connection.Close();
|
||||
}
|
||||
}
|
||||
|
||||
// INSERT-or-UPDATE for a post arriving from a Tumblr text-file export.
|
||||
@@ -2664,7 +2626,7 @@ namespace URLNotesGrabberCORE
|
||||
{
|
||||
minRetrySeconds = 0;
|
||||
if (!_usePool || _overrideKey != null) return false;
|
||||
if (_keys.Count <= 1) return true;
|
||||
if (_keys.Count == 0) return false;
|
||||
|
||||
var now = DateTimeOffset.UtcNow.ToUnixTimeSeconds();
|
||||
bool all = true;
|
||||
|
||||
+210
-93
@@ -5,7 +5,6 @@ using Microsoft.Extensions.Configuration;
|
||||
using System.Configuration;
|
||||
using System.Threading;
|
||||
using Microsoft.Extensions.Diagnostics.Latency;
|
||||
using static System.Runtime.InteropServices.JavaScript.JSType;
|
||||
using System.Text.RegularExpressions;
|
||||
|
||||
namespace URLNotesGrabberCORE
|
||||
@@ -13,12 +12,27 @@ namespace URLNotesGrabberCORE
|
||||
internal class Program
|
||||
{
|
||||
|
||||
static void Main(string[] args)
|
||||
static int Main(string[] args)
|
||||
{
|
||||
// Reset console color on exit (including Ctrl+C)
|
||||
Console.CancelKeyPress += (s, e) => Console.ResetColor();
|
||||
AppDomain.CurrentDomain.ProcessExit += (s, e) => Console.ResetColor();
|
||||
|
||||
try
|
||||
{
|
||||
return Run(args);
|
||||
}
|
||||
catch (Exception ex)
|
||||
{
|
||||
// Turn configuration errors / unguarded indexers into a quiet, deterministic exit code.
|
||||
Console.Error.WriteLine(ex.Message);
|
||||
return 1;
|
||||
}
|
||||
}
|
||||
|
||||
static int Run(string[] args)
|
||||
{
|
||||
int exitCode = 0;
|
||||
IConfiguration config = new ConfigurationBuilder()
|
||||
.SetBasePath(Directory.GetCurrentDirectory())
|
||||
.AddJsonFile("appsettings.json", optional: true, reloadOnChange: false)
|
||||
@@ -33,31 +47,34 @@ namespace URLNotesGrabberCORE
|
||||
List<string> filteredArgs = new List<string>();
|
||||
for (int i = 0; i < args.Length; i++)
|
||||
{
|
||||
if (string.Equals(args[i], "-force", StringComparison.OrdinalIgnoreCase) ||
|
||||
string.Equals(args[i], "--force", StringComparison.OrdinalIgnoreCase))
|
||||
if (args[i] == "--")
|
||||
{
|
||||
// POSIX end-of-options: everything after is a literal operand.
|
||||
for (int j = i + 1; j < args.Length; j++) filteredArgs.Add(args[j]);
|
||||
break;
|
||||
}
|
||||
|
||||
if (string.Equals(args[i], "--force", StringComparison.OrdinalIgnoreCase))
|
||||
{
|
||||
forceIgnoreCooldown = true;
|
||||
continue;
|
||||
}
|
||||
|
||||
if (string.Equals(args[i], "-api3", StringComparison.OrdinalIgnoreCase) ||
|
||||
string.Equals(args[i], "--api3", StringComparison.OrdinalIgnoreCase))
|
||||
if (string.Equals(args[i], "--api3", StringComparison.OrdinalIgnoreCase))
|
||||
{
|
||||
apiSectionName = "TumblrApi3";
|
||||
apiExplicitlySet = true;
|
||||
continue;
|
||||
}
|
||||
|
||||
if (string.Equals(args[i], "-api4", StringComparison.OrdinalIgnoreCase) ||
|
||||
string.Equals(args[i], "--api4", StringComparison.OrdinalIgnoreCase))
|
||||
if (string.Equals(args[i], "--api4", StringComparison.OrdinalIgnoreCase))
|
||||
{
|
||||
apiSectionName = "TumblrApi4";
|
||||
apiExplicitlySet = true;
|
||||
continue;
|
||||
}
|
||||
|
||||
if (string.Equals(args[i], "-api", StringComparison.OrdinalIgnoreCase) ||
|
||||
string.Equals(args[i], "--api", StringComparison.OrdinalIgnoreCase))
|
||||
if (string.Equals(args[i], "--api", StringComparison.OrdinalIgnoreCase))
|
||||
{
|
||||
if (i + 1 < args.Length && !string.IsNullOrWhiteSpace(args[i + 1]))
|
||||
{
|
||||
@@ -67,13 +84,12 @@ namespace URLNotesGrabberCORE
|
||||
}
|
||||
else
|
||||
{
|
||||
Console.WriteLine("--Missing API section after -api/--api. Using default TumblrApi.--");
|
||||
Console.WriteLine("--Missing API section after --api. Using default TumblrApi.--");
|
||||
}
|
||||
continue;
|
||||
}
|
||||
|
||||
if (string.Equals(args[i], "-start", StringComparison.OrdinalIgnoreCase) ||
|
||||
string.Equals(args[i], "--start", StringComparison.OrdinalIgnoreCase))
|
||||
if (string.Equals(args[i], "--start", StringComparison.OrdinalIgnoreCase))
|
||||
{
|
||||
if (i + 1 < args.Length && !string.IsNullOrWhiteSpace(args[i + 1]))
|
||||
{
|
||||
@@ -82,7 +98,7 @@ namespace URLNotesGrabberCORE
|
||||
}
|
||||
else
|
||||
{
|
||||
Console.WriteLine("--Missing blog name after -start/--start. Ignoring.--");
|
||||
Console.WriteLine("--Missing blog name after --start. Ignoring.--");
|
||||
}
|
||||
continue;
|
||||
}
|
||||
@@ -123,7 +139,10 @@ namespace URLNotesGrabberCORE
|
||||
Console.SetOut(dualLogger);
|
||||
}
|
||||
|
||||
List<string> contains = settings.GetValue<string>("ContainsList").Split(',').ToList();
|
||||
string? containsListSetting = settings.GetValue<string>("ContainsList");
|
||||
if (string.IsNullOrEmpty(containsListSetting))
|
||||
throw new InvalidOperationException("ContainsList is not configured in appsettings.json");
|
||||
List<string> contains = containsListSetting.Split(',').ToList();
|
||||
bool logTraversalRecordImports = settings.GetValue("LogTraversalRecordImports", false);
|
||||
|
||||
if (args.Length == 0) //Traverse folder structure to add posts and thus blogs to DB
|
||||
@@ -147,55 +166,23 @@ namespace URLNotesGrabberCORE
|
||||
switch (args[0])
|
||||
{
|
||||
case "-?":
|
||||
Console.WriteLine("\t Parse .txt files to find blogs");
|
||||
|
||||
Console.WriteLine("-?\t Usage help");
|
||||
|
||||
Console.WriteLine("-parse\t Parse .txt files with specified blogname");
|
||||
|
||||
Console.WriteLine("-test\t Calls API for given blogname and postID");
|
||||
|
||||
Console.WriteLine("-posts\t For each Post in DB, write blogname to file");
|
||||
|
||||
Console.WriteLine("-blogs\t For each Blog in DB, write blogname to file");
|
||||
|
||||
Console.WriteLine("-collect\t For each Post in DB, hit API to collect Notes. Optional datetime parameter to filter by NotesGatheredDateTime");
|
||||
|
||||
Console.WriteLine("-blogsR\t For each Note that is a REPLY, write blogname to file ");
|
||||
|
||||
Console.WriteLine("-blogsO\t For each Blog in DB, write blogname to file, but limit via a passed start and stop range ");
|
||||
|
||||
Console.WriteLine("-replies\t Fetch and update missing reply text for all replies in database");
|
||||
|
||||
Console.WriteLine("-likes\t Fetch likes: initial backfill for new blogs, incremental refresh for blogs past cooldown. Optional blog name forces single-blog run.");
|
||||
|
||||
Console.WriteLine("-force\t (with -likes) Ignore cooldown and refresh every fully-backfilled blog");
|
||||
|
||||
Console.WriteLine("-urldump\t Scan all posts' text columns and extract suspected URLs to configured file");
|
||||
|
||||
Console.WriteLine("-api3\t Use TumblrApi3 settings from appsettings.json");
|
||||
|
||||
Console.WriteLine("-api4\t Use TumblrApi4 settings from appsettings.json");
|
||||
|
||||
Console.WriteLine("-start [blogname]\t Start traversal alphabetically at this blog name");
|
||||
|
||||
Console.WriteLine("-api [section]\t Use a specific API settings section from appsettings.json (e.g. TumblrApi3)");
|
||||
|
||||
Console.WriteLine("-ingest [blogname]\t Ingest Tumblr .txt exports from appSettings:PathTTRoot into TL.db (all blogs, or single blog if name given)");
|
||||
|
||||
Console.WriteLine("-output\t Export posts from TL.db back to .txt files in each blog's TTFolderPath");
|
||||
|
||||
Console.WriteLine("-correct [bakPath]\t Dry-run: report multi-line field updates available from a BAK directory");
|
||||
|
||||
Console.WriteLine("-correct -apply [bakPath]\t Apply BAK-file corrections to matching posts (prompts yes/no)");
|
||||
|
||||
Console.WriteLine("-updatepaths [rootPath]\t Read .tumblr/.tmblrpriv metadata from <root>\\Index and set Blogs.TTFolderPath");
|
||||
|
||||
Console.WriteLine("-importposts [path-to-posts.db]\t One-time migration: copy legacy ThreeTxtFileHelper posts.db rows into TL.db");
|
||||
|
||||
case "-h":
|
||||
case "--help":
|
||||
PrintHelp();
|
||||
break;
|
||||
|
||||
case "-parse":
|
||||
|
||||
case "-V":
|
||||
case "--version":
|
||||
Console.WriteLine(System.Reflection.Assembly.GetExecutingAssembly().GetName().Version?.ToString() ?? "unknown");
|
||||
break;
|
||||
|
||||
case "--parse":
|
||||
if (args.Length < 2)
|
||||
{
|
||||
Console.WriteLine("Usage: --parse <blogname>");
|
||||
exitCode = 2;
|
||||
break;
|
||||
}
|
||||
string blogNameToParse = args[1];
|
||||
int postsAdded = 0;
|
||||
try
|
||||
@@ -210,29 +197,31 @@ namespace URLNotesGrabberCORE
|
||||
Console.WriteLine($"Total posts added: {postsAdded}");
|
||||
break;
|
||||
|
||||
case "-test":
|
||||
case "--test":
|
||||
Console.WriteLine("Test command not implemented");
|
||||
break;
|
||||
|
||||
case "-post":
|
||||
case "--post":
|
||||
TraverseDirectoryForCorruption(settings.GetValue<string>("PathInput"), settings.GetValue<string>("PathOutputBlogs"), contains);
|
||||
break;
|
||||
|
||||
case "-posts": //write post's blogs to file
|
||||
case "--posts": //write post's blogs to file
|
||||
WritePostBlogsToFile(settings.GetValue<string>("PathOutputPosts"));
|
||||
break;
|
||||
|
||||
case "-blogs": //write blogs to file
|
||||
case "--blogs": //write blogs to file
|
||||
WriteBlogsToFile(settings.GetValue<string>("PathOutputBlogs"));
|
||||
break;
|
||||
|
||||
case "-collect": //collect notes from all posts
|
||||
case "--collect": //collect notes from all posts
|
||||
bool withoutNotesOnly = true;
|
||||
DateTime? beforeDate = DateTime.Now;
|
||||
bool explicitDateSupplied = false;
|
||||
|
||||
if (args.Length < 2)
|
||||
{
|
||||
Console.WriteLine("--Expected WITHOUTNOTESONLY (0, 1) [OPTIONAL: BEFOREDATE]--");
|
||||
exitCode = 2;
|
||||
break;
|
||||
}
|
||||
|
||||
@@ -259,26 +248,50 @@ namespace URLNotesGrabberCORE
|
||||
if (DateTime.TryParse(args[2], out DateTime parsedDate))
|
||||
{
|
||||
beforeDate = parsedDate;
|
||||
explicitDateSupplied = true;
|
||||
Console.WriteLine($"Filter: Collecting notes for posts with NotesGatheredDateTime < {beforeDate}");
|
||||
}
|
||||
else
|
||||
{
|
||||
Console.WriteLine($"ERROR: Invalid date format '{args[2]}'");
|
||||
exitCode = 2;
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
CollectNotes(settings.GetValue<string>("PathOutput"), withoutNotesOnly, beforeDate).GetAwaiter().GetResult();
|
||||
// Mode 0 (full re-check) with no explicit date is a *managed* run: freeze the cutoff and
|
||||
// persist it so an interrupted run resumes against the same cutoff and a completed run stops
|
||||
// instead of restarting. Mode 1 and explicit-date runs keep their existing behavior.
|
||||
bool managedCollectRun = false;
|
||||
if (!withoutNotesOnly && !explicitDateSupplied)
|
||||
{
|
||||
DataAccess.EnsureCollectRunStateTableExists();
|
||||
var runState = DataAccess.GetCollectRunState();
|
||||
if (runState != null && !runState.Value.complete)
|
||||
{
|
||||
beforeDate = DateTimeOffset.FromUnixTimeSeconds(runState.Value.cutoff).LocalDateTime;
|
||||
Console.WriteLine($"Resuming interrupted full re-check (cutoff = {beforeDate})");
|
||||
}
|
||||
else
|
||||
{
|
||||
beforeDate = DateTime.Now;
|
||||
DataAccess.BeginCollectRun(new DateTimeOffset(beforeDate.Value).ToUnixTimeSeconds());
|
||||
Console.WriteLine($"Starting new full re-check run (cutoff = {beforeDate})");
|
||||
}
|
||||
managedCollectRun = true;
|
||||
}
|
||||
|
||||
CollectNotes(settings.GetValue<string>("PathOutput"), withoutNotesOnly, beforeDate, managedCollectRun).GetAwaiter().GetResult();
|
||||
break;
|
||||
|
||||
case "-blogsR": //collect notes from all posts
|
||||
case "--blogsR": //collect notes from all posts
|
||||
WriteBlogsToFile(settings.GetValue<string>("PathOutputBlogs"), true);
|
||||
break;
|
||||
|
||||
case "-blogsO": //collect notes from all posts
|
||||
case "--blogsO": //collect notes from all posts
|
||||
int from = 1, to = 999999, top = 100;
|
||||
|
||||
if (args[1] is not null && args[2] is not null && args[3] is not null)
|
||||
if (args.Length >= 4 && args[1] is not null && args[2] is not null && args[3] is not null)
|
||||
{
|
||||
from = int.Parse(args[1]);
|
||||
to = int.Parse(args[2]);
|
||||
@@ -287,14 +300,16 @@ namespace URLNotesGrabberCORE
|
||||
else
|
||||
{
|
||||
Console.WriteLine("--Expected FROM TO--");
|
||||
exitCode = 2;
|
||||
break;
|
||||
}
|
||||
WriteBlogsToFile(settings.GetValue<string>("PathOutputBlogs"), false, from, to, top);
|
||||
break;
|
||||
|
||||
case "-bop": //collect notes from all posts
|
||||
case "--bop": //collect notes from all posts
|
||||
from = 1; to = 999999; top = 100;
|
||||
|
||||
if (args[1] is not null && args[2] is not null && args[3] is not null)
|
||||
if (args.Length >= 4 && args[1] is not null && args[2] is not null && args[3] is not null)
|
||||
{
|
||||
from = int.Parse(args[1]);
|
||||
to = int.Parse(args[2]);
|
||||
@@ -303,62 +318,70 @@ namespace URLNotesGrabberCORE
|
||||
else
|
||||
{
|
||||
Console.WriteLine("--Expected FROM TO--");
|
||||
exitCode = 2;
|
||||
break;
|
||||
}
|
||||
WriteBlogsToFileAll(settings.GetValue<string>("PathOutputBlogs"), false, from, to, top);
|
||||
break;
|
||||
|
||||
case "-replies": //update reply text
|
||||
case "--replies": //update reply text
|
||||
CollectMissingReplyText().GetAwaiter().GetResult();
|
||||
break;
|
||||
|
||||
case "-likes":
|
||||
case "--likes":
|
||||
string likeBlog = args.Length > 1 ? args[1] : null;
|
||||
int cooldownDays = settings.GetValue("LikesRefreshCooldownDays", 7);
|
||||
CollectLikes(likeBlog, contains, cooldownDays, forceIgnoreCooldown).GetAwaiter().GetResult();
|
||||
break;
|
||||
|
||||
case "-urldump":
|
||||
case "--urldump":
|
||||
DumpUrls(settings.GetValue<string>("PathOutputUrls"));
|
||||
break;
|
||||
|
||||
case "-ingest":
|
||||
IngestMode.Run(config, args.Skip(1).ToArray());
|
||||
case "--ingest":
|
||||
exitCode = IngestMode.Run(config, args.Skip(1).ToArray());
|
||||
break;
|
||||
|
||||
case "-output":
|
||||
OutputMode.Run(config);
|
||||
case "--output":
|
||||
exitCode = OutputMode.Run(config);
|
||||
break;
|
||||
|
||||
case "-correct":
|
||||
case "--revert":
|
||||
exitCode = RevertMode.Run(config, args.Length > 1 ? args[1] : null);
|
||||
break;
|
||||
|
||||
case "--correct":
|
||||
{
|
||||
bool applyChanges = args.Skip(1).Any(a => string.Equals(a, "-apply", StringComparison.OrdinalIgnoreCase) || string.Equals(a, "--apply", StringComparison.OrdinalIgnoreCase));
|
||||
bool applyChanges = args.Skip(1).Any(a => string.Equals(a, "--apply", StringComparison.OrdinalIgnoreCase));
|
||||
var correctArgs = args.Skip(1)
|
||||
.Where(a => !string.Equals(a, "-apply", StringComparison.OrdinalIgnoreCase) && !string.Equals(a, "--apply", StringComparison.OrdinalIgnoreCase))
|
||||
.Where(a => !string.Equals(a, "--apply", StringComparison.OrdinalIgnoreCase))
|
||||
.ToArray();
|
||||
CorrectMode.Run(config, correctArgs, applyChanges);
|
||||
exitCode = CorrectMode.Run(config, correctArgs, applyChanges);
|
||||
break;
|
||||
}
|
||||
|
||||
case "-updatepaths":
|
||||
case "--updatepaths":
|
||||
{
|
||||
string rootPath = args.Length > 1 ? args[1] : (settings.GetValue<string>("PathTTRoot") ?? settings.GetValue<string>("PathInput") ?? string.Empty);
|
||||
UpdateBlogPathsRunner.Run(rootPath);
|
||||
exitCode = UpdateBlogPathsRunner.Run(rootPath);
|
||||
break;
|
||||
}
|
||||
|
||||
case "-importposts":
|
||||
case "--importposts":
|
||||
{
|
||||
if (args.Length < 2)
|
||||
{
|
||||
Console.WriteLine("Usage: -importposts <path-to-legacy-posts.db>");
|
||||
Console.WriteLine("Usage: --importposts <path-to-legacy-posts.db>");
|
||||
exitCode = 2;
|
||||
break;
|
||||
}
|
||||
LegacyPostsDbImporter.Run(args[1]);
|
||||
exitCode = LegacyPostsDbImporter.Run(args[1]);
|
||||
break;
|
||||
}
|
||||
|
||||
default:
|
||||
Console.WriteLine("** Unknown Command ** " + args[0]);
|
||||
exitCode = 2;
|
||||
break;
|
||||
}
|
||||
}
|
||||
@@ -367,6 +390,69 @@ namespace URLNotesGrabberCORE
|
||||
|
||||
System.Console.WriteLine("<fin>:/");
|
||||
//System.Console.ReadKey();
|
||||
return exitCode;
|
||||
}
|
||||
|
||||
static void PrintHelp()
|
||||
{
|
||||
Console.WriteLine("\t Parse .txt files to find blogs");
|
||||
|
||||
Console.WriteLine("-?, -h, --help\t Usage help");
|
||||
|
||||
Console.WriteLine("-V, --version\t Print the application version");
|
||||
|
||||
Console.WriteLine("--\t End of options: treat every following token as a literal operand");
|
||||
|
||||
Console.WriteLine("--parse\t Parse .txt files with specified blogname");
|
||||
|
||||
Console.WriteLine("--test\t Calls API for given blogname and postID");
|
||||
|
||||
Console.WriteLine("--post\t Traverse the input directory tree checking .txt files for corruption");
|
||||
|
||||
Console.WriteLine("--posts\t For each Post in DB, write blogname to file");
|
||||
|
||||
Console.WriteLine("--blogs\t For each Blog in DB, write blogname to file");
|
||||
|
||||
Console.WriteLine("--collect [0|1] [datetime]\t Collect Notes from API. 1=only posts without notes. 0=full re-check of all posts: a single resumable pass (interrupt & relaunch to resume; stops when complete, retrigger for a new pass). Optional datetime overrides the cutoff and runs as a one-off (bypasses resume tracking).");
|
||||
|
||||
Console.WriteLine("--blogsR\t For each Note that is a REPLY, write blogname to file ");
|
||||
|
||||
Console.WriteLine("--blogsO\t For each Blog in DB, write blogname to file, but limit via a passed start and stop range ");
|
||||
|
||||
Console.WriteLine("--bop [from] [to] [top]\t Write ALL blog names to file, limited by FROM TO TOP range arguments");
|
||||
|
||||
Console.WriteLine("--replies\t Fetch and update missing reply text for all replies in database");
|
||||
|
||||
Console.WriteLine("--likes\t Fetch likes: initial backfill for new blogs, incremental refresh for blogs past cooldown. Optional blog name forces single-blog run.");
|
||||
|
||||
Console.WriteLine("--force\t (with --likes) Ignore cooldown and refresh every fully-backfilled blog");
|
||||
|
||||
Console.WriteLine("--urldump\t Scan all posts' text columns and extract suspected URLs to configured file");
|
||||
|
||||
Console.WriteLine("--api3\t Use TumblrApi3 settings from appsettings.json");
|
||||
|
||||
Console.WriteLine("--api4\t Use TumblrApi4 settings from appsettings.json");
|
||||
|
||||
Console.WriteLine("--start [blogname]\t Start traversal alphabetically at this blog name");
|
||||
|
||||
Console.WriteLine("--api [section]\t Use a specific API settings section from appsettings.json (e.g. TumblrApi3)");
|
||||
|
||||
Console.WriteLine("--ingest [blogname]\t Ingest Tumblr .txt exports from appSettings:PathTTRoot into TL.db (all blogs, or single blog if name given)");
|
||||
|
||||
Console.WriteLine("--output\t Export posts from TL.db back to .txt files in each blog's TTFolderPath");
|
||||
|
||||
Console.WriteLine("--revert [blogname]\t Recursively scan the PathInput tree and restore *.bak back to *.txt (current .txt saved as next-free .bkN); optional blogname filters by path substring");
|
||||
|
||||
Console.WriteLine("--correct [bakPath]\t Dry-run: report multi-line field updates available from a BAK directory");
|
||||
|
||||
Console.WriteLine("--correct --apply [bakPath]\t Apply BAK-file corrections to matching posts (prompts yes/no)");
|
||||
|
||||
Console.WriteLine("--updatepaths [rootPath]\t Read .tumblr/.tmblrpriv metadata from <root>\\Index and set Blogs.TTFolderPath");
|
||||
|
||||
Console.WriteLine("--importposts [path-to-posts.db]\t One-time migration: copy legacy ThreeTxtFileHelper posts.db rows into TL.db");
|
||||
|
||||
Console.WriteLine();
|
||||
Console.WriteLine("Exit status: 0 = success; 1 = unexpected error; 2 = usage error (unknown command or bad/missing arguments)");
|
||||
}
|
||||
|
||||
static void WritePostBlogsToFile(string outPath)
|
||||
@@ -1182,10 +1268,15 @@ if (shouldInsert)
|
||||
return "UNKNOWN";
|
||||
}
|
||||
|
||||
static async Task CollectNotes(string outPath, bool withoutNotesOnly = true, DateTime? beforeDate = null)
|
||||
static async Task CollectNotes(string outPath, bool withoutNotesOnly = true, DateTime? beforeDate = null, bool managedRun = false)
|
||||
{
|
||||
List<Tuple<string, long, long, long>> posts = DataAccess.GetPosts(withoutNotesOnly, beforeDate);
|
||||
|
||||
// Posts attempted (with a definitive, non-throttle result) during *this* process. Guarantees a single
|
||||
// attempt pass: once every remaining post has been attempted, the loop stops instead of spinning on a
|
||||
// post that keeps returning FAILURE/UNKNOWN. Successful/NotFound posts drop out via the DB filter anyway.
|
||||
HashSet<(string, long)> attempted = new HashSet<(string, long)>();
|
||||
|
||||
RateLimiter limiter = new SlidingWindowRateLimiter(new SlidingWindowRateLimiterOptions
|
||||
{
|
||||
PermitLimit = 300,
|
||||
@@ -1202,9 +1293,17 @@ if (shouldInsert)
|
||||
{
|
||||
while (posts.Count > 0)
|
||||
{
|
||||
// First post not yet attempted this process. If all remaining have been attempted, the pass
|
||||
// is done (the stragglers returned FAILURE/UNKNOWN) — stop rather than loop forever.
|
||||
var post = posts.FirstOrDefault(p => !attempted.Contains((p.Item1, p.Item2)));
|
||||
if (post == null)
|
||||
{
|
||||
Console.WriteLine("All remaining posts have been attempted this run; ending pass.");
|
||||
break;
|
||||
}
|
||||
|
||||
ApiKeyPool.SleepUntilAnyAvailable(30);
|
||||
|
||||
var post = posts[0]; // Process the first post in the list
|
||||
string status;
|
||||
|
||||
using RateLimitLease lease = limiter.AttemptAcquire(1);
|
||||
@@ -1216,20 +1315,31 @@ if (shouldInsert)
|
||||
else
|
||||
{
|
||||
Console.WriteLine("!@@@@@@@ - Rate Limited Exceeded: No Lease Available");
|
||||
return;
|
||||
return; // throttle: abort without completing the run so a later launch resumes
|
||||
}
|
||||
|
||||
if (status == "Success")
|
||||
{
|
||||
attempted.Add((post.Item1, post.Item2));
|
||||
DataAccess.UpdatePostMarkNotesCollected(post.Item1, post.Item2);
|
||||
}
|
||||
else if (status == "NotFound")
|
||||
{
|
||||
attempted.Add((post.Item1, post.Item2));
|
||||
Console.WriteLine("GrabNotes Result: NotFound");
|
||||
DataAccess.UpdatePostMarkNotFound(post.Item1, post.Item2);
|
||||
}
|
||||
else if (status == "TooManyRequests")
|
||||
{
|
||||
// Throttle, not a real per-post failure: don't consume this post's single attempt.
|
||||
// Abort the pass without completing so a later launch resumes against the same cutoff.
|
||||
Console.WriteLine("GrabNotes Result: TooManyRequests - pausing run; relaunch to resume.");
|
||||
return;
|
||||
}
|
||||
else
|
||||
{
|
||||
// FAILURE / UNKNOWN: count as attempted so the pass can finish instead of retrying forever.
|
||||
attempted.Add((post.Item1, post.Item2));
|
||||
Console.WriteLine("GrabNotes Result: " + status);
|
||||
}
|
||||
|
||||
@@ -1237,6 +1347,13 @@ if (shouldInsert)
|
||||
posts = DataAccess.GetPosts(withoutNotesOnly, beforeDate);
|
||||
}
|
||||
}
|
||||
|
||||
// Reached only when the pass finished naturally (worklist drained or all stragglers attempted).
|
||||
if (managedRun)
|
||||
{
|
||||
DataAccess.CompleteCollectRun();
|
||||
Console.WriteLine("Full re-check run complete.");
|
||||
}
|
||||
}
|
||||
catch (Exception ex)
|
||||
{
|
||||
|
||||
@@ -2,7 +2,7 @@
|
||||
"profiles": {
|
||||
"URLNotesGrabberCORE": {
|
||||
"commandName": "Project",
|
||||
"commandLineArgs": "-collect 1 -api4"
|
||||
"commandLineArgs": "--collect 1 --api4"
|
||||
}
|
||||
}
|
||||
}
|
||||
@@ -0,0 +1,110 @@
|
||||
using Microsoft.Extensions.Configuration;
|
||||
|
||||
namespace URLNotesGrabberCORE
|
||||
{
|
||||
// Inverse of OutputMode. Recursively walks the PathInput tree (the same directory tree the
|
||||
// no-parameter run uses) and restores every *.bak back to its *.txt, first preserving the
|
||||
// current *.txt as the next-free *.bkN. Consumes the *.bak (File.Move). Filesystem-only;
|
||||
// does not read the DB. An optional blogname argument filters by path substring.
|
||||
public static class RevertMode
|
||||
{
|
||||
public static int Run(IConfiguration config, string? blogFilter = null)
|
||||
{
|
||||
string? root = config["appSettings:PathInput"];
|
||||
if (string.IsNullOrWhiteSpace(root) || !Directory.Exists(root))
|
||||
{
|
||||
Console.WriteLine($"PathInput is not set or does not exist: '{root}'. Nothing to revert.");
|
||||
return 0;
|
||||
}
|
||||
|
||||
Console.WriteLine($"Searching for .bak files under: {root}");
|
||||
|
||||
// Recursively collect every *.bak, optionally filtered by path substring (blogname).
|
||||
var bakFiles = EnumerateBakFiles(root)
|
||||
.Where(f => string.IsNullOrWhiteSpace(blogFilter)
|
||||
|| f.IndexOf(blogFilter, StringComparison.OrdinalIgnoreCase) >= 0)
|
||||
.ToList();
|
||||
|
||||
if (bakFiles.Count == 0)
|
||||
{
|
||||
Console.WriteLine("No .bak files found. Nothing to revert.");
|
||||
return 0;
|
||||
}
|
||||
|
||||
Console.Write($"WARNING: This will restore {bakFiles.Count} .bak file(s) over their .txt files. " +
|
||||
$"Current .txt files are preserved as the next-free .bkN. Continue? (yes/no): ");
|
||||
string? response = Console.ReadLine();
|
||||
if (string.IsNullOrWhiteSpace(response) || !response.Equals("yes", StringComparison.OrdinalIgnoreCase))
|
||||
{
|
||||
Console.WriteLine("Operation cancelled.");
|
||||
return 0;
|
||||
}
|
||||
|
||||
int restored = 0, backedUp = 0;
|
||||
foreach (var bakFile in bakFiles)
|
||||
{
|
||||
try
|
||||
{
|
||||
string txtPath = Path.ChangeExtension(bakFile, ".txt");
|
||||
|
||||
if (File.Exists(txtPath))
|
||||
{
|
||||
string bkPath = NextFreeBkPath(txtPath);
|
||||
File.Move(txtPath, bkPath);
|
||||
backedUp++;
|
||||
Console.WriteLine($" Backed up {Path.GetFileName(txtPath)} -> {Path.GetFileName(bkPath)}");
|
||||
}
|
||||
|
||||
File.Move(bakFile, txtPath);
|
||||
restored++;
|
||||
Console.WriteLine($" Restored {bakFile} -> {Path.GetFileName(txtPath)}");
|
||||
}
|
||||
catch (Exception ex)
|
||||
{
|
||||
Console.WriteLine($" Error reverting {bakFile}: {ex.Message}");
|
||||
}
|
||||
}
|
||||
|
||||
Console.WriteLine($"\nRevert mode complete. Restored {restored} file(s); backed up {backedUp} current .txt file(s).");
|
||||
return 0;
|
||||
}
|
||||
|
||||
// Recursively yields every *.bak path under root. Per-directory try/catch so an
|
||||
// inaccessible folder doesn't abort the whole walk (mirrors TraverseDirectory).
|
||||
private static IEnumerable<string> EnumerateBakFiles(string path)
|
||||
{
|
||||
string[] subDirs;
|
||||
try { subDirs = Directory.GetDirectories(path); }
|
||||
catch (Exception ex)
|
||||
{
|
||||
Console.WriteLine($" Skipping '{path}': {ex.Message}");
|
||||
yield break;
|
||||
}
|
||||
|
||||
foreach (var dir in subDirs)
|
||||
foreach (var bak in EnumerateBakFiles(dir))
|
||||
yield return bak;
|
||||
|
||||
string[] bakFiles;
|
||||
try { bakFiles = Directory.GetFiles(path, "*.bak"); }
|
||||
catch (Exception ex)
|
||||
{
|
||||
Console.WriteLine($" Skipping files in '{path}': {ex.Message}");
|
||||
yield break;
|
||||
}
|
||||
|
||||
foreach (var bak in bakFiles)
|
||||
yield return bak;
|
||||
}
|
||||
|
||||
// Returns the lowest unused .bkN path for a given .txt file (.bk1, .bk2, ...).
|
||||
private static string NextFreeBkPath(string txtFile)
|
||||
{
|
||||
for (int n = 1; ; n++)
|
||||
{
|
||||
string candidate = Path.ChangeExtension(txtFile, $".bk{n}");
|
||||
if (!File.Exists(candidate)) return candidate;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
@@ -1,52 +0,0 @@
|
||||
@echo off
|
||||
REM Batch file to run URLNotesGrabberCORE 500 times in a loop
|
||||
|
||||
setlocal enabledelayedexpansion
|
||||
|
||||
REM Set the path to the executable
|
||||
REM Update this path if your executable is in a different location
|
||||
set APP_PATH=URLNotesGrabberCORE.exe
|
||||
|
||||
REM Check if the executable exists
|
||||
if not exist "%APP_PATH%" (
|
||||
echo Error: %APP_PATH% not found in the current directory.
|
||||
echo Please ensure the executable is in the same directory as this batch file,
|
||||
echo or update the APP_PATH variable with the correct path.
|
||||
pause
|
||||
exit /b 1
|
||||
)
|
||||
|
||||
REM Loop counter
|
||||
set ITERATIONS=500
|
||||
set COUNTER=0
|
||||
|
||||
echo Starting to run %APP_PATH% %ITERATIONS% times...
|
||||
echo.
|
||||
|
||||
:LOOP
|
||||
set /a COUNTER+=1
|
||||
echo [%COUNTER%/%ITERATIONS%] Running iteration %COUNTER%...
|
||||
echo Started at: %date% %time%
|
||||
|
||||
REM Run the application with -replies option
|
||||
call "%APP_PATH%" -replies
|
||||
|
||||
REM Check if the application ran successfully
|
||||
if errorlevel 1 (
|
||||
echo Warning: Application exited with error code !ERRORLEVEL! on iteration %COUNTER%
|
||||
) else (
|
||||
echo Iteration %COUNTER% completed successfully.
|
||||
)
|
||||
|
||||
echo Completed at: %date% %time%
|
||||
echo.
|
||||
|
||||
REM Check if we've reached 500 iterations
|
||||
if %COUNTER% lss %ITERATIONS% (
|
||||
goto LOOP
|
||||
)
|
||||
|
||||
echo.
|
||||
echo Completed all %ITERATIONS% iterations!
|
||||
echo.
|
||||
pause
|
||||
@@ -0,0 +1,199 @@
|
||||
-- ============================================================================
|
||||
-- verify-db-schema.sql
|
||||
--
|
||||
-- Purpose: Verify that a TL.db (e.g. a restored backup) has every column the
|
||||
-- current URLNotesGrabberCORE code expects. The app has NO startup
|
||||
-- migration: missing columns only get added when specific modes run,
|
||||
-- and a referenced-but-missing column causes a "no such column" crash.
|
||||
--
|
||||
-- How to use (DB Browser for SQLite):
|
||||
-- 1. File > Open Database -> pick the restored backup.
|
||||
-- 2. Execute SQL tab. Run SECTION 1 (it is read-only).
|
||||
-- * Zero rows from every query = schema is fully aligned, you're done.
|
||||
-- * Rows in "MISSING COLUMNS" = copy the run_this_to_fix text.
|
||||
-- 3. If columns are missing: KEEP A COPY OF THE BACKUP FIRST, then go to
|
||||
-- SECTION 2, uncomment ONLY the ALTER lines that match the report, and run.
|
||||
-- 4. Re-run SECTION 1 to confirm zero rows.
|
||||
--
|
||||
-- This script never UPDATEs/DELETEs/DROPs. In particular it deliberately does
|
||||
-- NOT replicate the likes-reset that the app's -likes migration performs
|
||||
-- (DataAccess.cs:375), so existing likes high-water marks are preserved.
|
||||
-- ============================================================================
|
||||
|
||||
|
||||
-- ============================================================================
|
||||
-- SECTION 1 -- VERIFICATION (read-only)
|
||||
-- ============================================================================
|
||||
|
||||
-- Expected schema for the current code version.
|
||||
-- alter_stmt is a runnable ALTER for additively-fixable columns; for base
|
||||
-- columns it is a 'MANUAL REVIEW' note (a missing base column means the backup
|
||||
-- predates the table's creation or is damaged -- do not blindly auto-add).
|
||||
WITH expected(tbl, col, alter_stmt) AS (
|
||||
VALUES
|
||||
-- Posts (base columns: manual review if missing)
|
||||
('Posts','BlogName', 'MANUAL REVIEW - base/PK column missing'),
|
||||
('Posts','PostID', 'MANUAL REVIEW - base/PK column missing'),
|
||||
('Posts','HasNotesGathered', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','reblogURL', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','NotFound', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','PostDate', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','NotesGatheredDateTime', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','HasImage', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','PostURL', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','Slug', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','ReblogKey', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','ReblogName', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','Summary', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','Quote', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','Body', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','Tags', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','Link', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','PhotoURL', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','PhotoCaption', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','DownloadedFiles', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','AudioCaption', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','Question', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','Answer', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','Title', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','ByLikes', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','RootBlogName', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','RootURL', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','DateModified', 'MANUAL REVIEW - base column missing'),
|
||||
('Posts','DateCreated', 'MANUAL REVIEW - base column missing'),
|
||||
-- Posts (additive migration column, auto-fixable)
|
||||
('Posts','PostType', 'ALTER TABLE Posts ADD COLUMN PostType TEXT;'),
|
||||
|
||||
-- Blogs (base columns: manual review if missing)
|
||||
('Blogs','BlogName', 'MANUAL REVIEW - base/PK column missing'),
|
||||
('Blogs','HasBeenOutput', 'MANUAL REVIEW - base column missing'),
|
||||
('Blogs','IsActive', 'MANUAL REVIEW - base column missing'),
|
||||
('Blogs','DateAdded', 'MANUAL REVIEW - base column missing'),
|
||||
('Blogs','ByLikes', 'MANUAL REVIEW - base column missing'),
|
||||
('Blogs','DateModified', 'MANUAL REVIEW - base column missing'),
|
||||
('Blogs','DateCreated', 'MANUAL REVIEW - base column missing'),
|
||||
-- Blogs (additive migration columns, auto-fixable)
|
||||
('Blogs','LikesPulled', 'ALTER TABLE Blogs ADD COLUMN LikesPulled INTEGER DEFAULT 0;'),
|
||||
('Blogs','LikesCursor', 'ALTER TABLE Blogs ADD COLUMN LikesCursor INTEGER DEFAULT 0;'),
|
||||
('Blogs','LikesNewestTimestamp', 'ALTER TABLE Blogs ADD COLUMN LikesNewestTimestamp INTEGER DEFAULT 0;'),
|
||||
('Blogs','LikesLastRefreshed', 'ALTER TABLE Blogs ADD COLUMN LikesLastRefreshed INTEGER DEFAULT 0;'),
|
||||
('Blogs','LikesLastNewCount', 'ALTER TABLE Blogs ADD COLUMN LikesLastNewCount INTEGER DEFAULT 0;'),
|
||||
('Blogs','TTFolderPath', 'ALTER TABLE Blogs ADD COLUMN TTFolderPath TEXT;'),
|
||||
|
||||
-- Notes (base columns: manual review if missing)
|
||||
('Notes','RootBlogName', 'MANUAL REVIEW - base/PK column missing'),
|
||||
('Notes','PostID', 'MANUAL REVIEW - base/PK column missing'),
|
||||
('Notes','NoteBlogName', 'MANUAL REVIEW - base/PK column missing'),
|
||||
('Notes','TimeStamp', 'MANUAL REVIEW - base/PK column missing'),
|
||||
('Notes','Type', 'MANUAL REVIEW - base/PK column missing'),
|
||||
('Notes','DatetimeCrawled', 'MANUAL REVIEW - base column missing'),
|
||||
('Notes','DateModified', 'MANUAL REVIEW - base column missing'),
|
||||
('Notes','DateCreated', 'MANUAL REVIEW - base column missing'),
|
||||
-- Notes (additive migration column, auto-fixable)
|
||||
('Notes','replyText', 'ALTER TABLE Notes ADD COLUMN replyText TEXT DEFAULT ''.'';'),
|
||||
|
||||
-- DailyAPICount (base columns)
|
||||
('DailyAPICount','Date', 'MANUAL REVIEW - base/PK column missing'),
|
||||
('DailyAPICount','APICount', 'MANUAL REVIEW - base column missing'),
|
||||
|
||||
-- ApiKeyPoolState (created at runtime by EnsureApiKeyPoolTables; auto-fixable by re-running app, but safe to add)
|
||||
('ApiKeyPoolState','KeyName', 'MANUAL REVIEW - run app once to auto-create ApiKeyPool tables'),
|
||||
('ApiKeyPoolState','RetryUntil', 'MANUAL REVIEW - run app once to auto-create ApiKeyPool tables'),
|
||||
('ApiKeyPoolMeta','Id', 'MANUAL REVIEW - run app once to auto-create ApiKeyPool tables'),
|
||||
('ApiKeyPoolMeta','LastIndex', 'MANUAL REVIEW - run app once to auto-create ApiKeyPool tables')
|
||||
),
|
||||
actual(tbl, col) AS (
|
||||
SELECT 'Posts', name FROM pragma_table_info('Posts')
|
||||
UNION ALL SELECT 'Blogs', name FROM pragma_table_info('Blogs')
|
||||
UNION ALL SELECT 'Notes', name FROM pragma_table_info('Notes')
|
||||
UNION ALL SELECT 'DailyAPICount', name FROM pragma_table_info('DailyAPICount')
|
||||
UNION ALL SELECT 'ApiKeyPoolState', name FROM pragma_table_info('ApiKeyPoolState')
|
||||
UNION ALL SELECT 'ApiKeyPoolMeta', name FROM pragma_table_info('ApiKeyPoolMeta')
|
||||
)
|
||||
|
||||
-- 1a. MISSING COLUMNS: columns the code needs that the DB does not have.
|
||||
-- Zero rows = good. Otherwise copy run_this_to_fix into SECTION 2.
|
||||
SELECT
|
||||
e.tbl AS table_name,
|
||||
e.col AS missing_column,
|
||||
e.alter_stmt AS run_this_to_fix
|
||||
FROM expected e
|
||||
LEFT JOIN actual a
|
||||
ON a.tbl = e.tbl AND lower(a.col) = lower(e.col)
|
||||
WHERE a.col IS NULL
|
||||
ORDER BY (e.alter_stmt LIKE 'ALTER%') DESC, e.tbl, e.col;
|
||||
|
||||
|
||||
-- 1b. MISSING TABLES: expected tables that don't exist at all in this DB.
|
||||
-- Zero rows = good.
|
||||
WITH expected_tables(tbl) AS (
|
||||
VALUES ('Posts'),('Blogs'),('Notes'),('DailyAPICount'),
|
||||
('ApiKeyPoolState'),('ApiKeyPoolMeta')
|
||||
)
|
||||
SELECT et.tbl AS missing_table
|
||||
FROM expected_tables et
|
||||
WHERE NOT EXISTS (
|
||||
SELECT 1 FROM sqlite_master
|
||||
WHERE type = 'table' AND lower(name) = lower(et.tbl)
|
||||
)
|
||||
ORDER BY et.tbl;
|
||||
|
||||
|
||||
-- 1c. EXTRA / UNEXPECTED COLUMNS: present in the DB but not in the expected
|
||||
-- list above. Informational only -- e.g. a NEWER backup, or a column this
|
||||
-- script's expected-list hasn't been updated for. Not an error by itself.
|
||||
WITH expected(tbl, col) AS (
|
||||
VALUES
|
||||
('Posts','BlogName'),('Posts','PostID'),('Posts','HasNotesGathered'),('Posts','reblogURL'),
|
||||
('Posts','NotFound'),('Posts','PostDate'),('Posts','NotesGatheredDateTime'),('Posts','HasImage'),
|
||||
('Posts','PostURL'),('Posts','Slug'),('Posts','ReblogKey'),('Posts','ReblogName'),('Posts','Summary'),
|
||||
('Posts','Quote'),('Posts','Body'),('Posts','Tags'),('Posts','Link'),('Posts','PhotoURL'),
|
||||
('Posts','PhotoCaption'),('Posts','DownloadedFiles'),('Posts','AudioCaption'),('Posts','Question'),
|
||||
('Posts','Answer'),('Posts','Title'),('Posts','ByLikes'),('Posts','RootBlogName'),('Posts','RootURL'),
|
||||
('Posts','DateModified'),('Posts','DateCreated'),('Posts','PostType'),
|
||||
('Blogs','BlogName'),('Blogs','HasBeenOutput'),('Blogs','IsActive'),('Blogs','DateAdded'),
|
||||
('Blogs','ByLikes'),('Blogs','DateModified'),('Blogs','DateCreated'),('Blogs','LikesPulled'),
|
||||
('Blogs','LikesCursor'),('Blogs','LikesNewestTimestamp'),('Blogs','LikesLastRefreshed'),
|
||||
('Blogs','LikesLastNewCount'),('Blogs','TTFolderPath'),
|
||||
('Notes','RootBlogName'),('Notes','PostID'),('Notes','NoteBlogName'),('Notes','TimeStamp'),
|
||||
('Notes','Type'),('Notes','DatetimeCrawled'),('Notes','DateModified'),('Notes','DateCreated'),
|
||||
('Notes','replyText'),
|
||||
('DailyAPICount','Date'),('DailyAPICount','APICount'),
|
||||
('ApiKeyPoolState','KeyName'),('ApiKeyPoolState','RetryUntil'),
|
||||
('ApiKeyPoolMeta','Id'),('ApiKeyPoolMeta','LastIndex')
|
||||
),
|
||||
actual(tbl, col) AS (
|
||||
SELECT 'Posts', name FROM pragma_table_info('Posts')
|
||||
UNION ALL SELECT 'Blogs', name FROM pragma_table_info('Blogs')
|
||||
UNION ALL SELECT 'Notes', name FROM pragma_table_info('Notes')
|
||||
UNION ALL SELECT 'DailyAPICount', name FROM pragma_table_info('DailyAPICount')
|
||||
UNION ALL SELECT 'ApiKeyPoolState', name FROM pragma_table_info('ApiKeyPoolState')
|
||||
UNION ALL SELECT 'ApiKeyPoolMeta', name FROM pragma_table_info('ApiKeyPoolMeta')
|
||||
)
|
||||
SELECT a.tbl AS table_name, a.col AS unexpected_column
|
||||
FROM actual a
|
||||
LEFT JOIN expected e
|
||||
ON e.tbl = a.tbl AND lower(e.col) = lower(a.col)
|
||||
WHERE e.col IS NULL
|
||||
ORDER BY a.tbl, a.col;
|
||||
|
||||
|
||||
-- ============================================================================
|
||||
-- SECTION 2 -- FIX (opt-in, additive only)
|
||||
--
|
||||
-- Run ONLY the lines that query 1a flagged with an ALTER statement.
|
||||
-- KEEP A COPY OF THE BACKUP FIRST. SQLite has no "ADD COLUMN IF NOT EXISTS",
|
||||
-- so running an ALTER for a column that already exists throws a harmless
|
||||
-- "duplicate column name" error and changes nothing -- just run the flagged
|
||||
-- subset. These are the 8 additive migration columns and nothing else; the
|
||||
-- likes high-water-mark reset is intentionally NOT included.
|
||||
-- ============================================================================
|
||||
|
||||
-- ALTER TABLE Posts ADD COLUMN PostType TEXT;
|
||||
-- ALTER TABLE Blogs ADD COLUMN LikesPulled INTEGER DEFAULT 0;
|
||||
-- ALTER TABLE Blogs ADD COLUMN LikesCursor INTEGER DEFAULT 0;
|
||||
-- ALTER TABLE Blogs ADD COLUMN LikesNewestTimestamp INTEGER DEFAULT 0;
|
||||
-- ALTER TABLE Blogs ADD COLUMN LikesLastRefreshed INTEGER DEFAULT 0;
|
||||
-- ALTER TABLE Blogs ADD COLUMN LikesLastNewCount INTEGER DEFAULT 0;
|
||||
-- ALTER TABLE Blogs ADD COLUMN TTFolderPath TEXT;
|
||||
-- ALTER TABLE Notes ADD COLUMN replyText TEXT DEFAULT '.';
|
||||
Reference in New Issue
Block a user