[#51] Address review: case-sensitive archive names, fix doc link and comments

- Compare archive names case-sensitively so the code matches the case-sensitive
  archives.name UNIQUE constraint and the name as it exists on the file system
  (drops the OrdinalIgnoreCase set).
- Fix the comparing-builds.md link to be same-folder relative.
- Clarify the schema comment and a test comment.

Author: SkowronskiAndrew

Analyzer/Resources/Init.sql

@@ -8,9 +8,10 @@ CREATE TABLE IF NOT EXISTS types
 -- Describes a unity archive that contains serialized files and other built content.
 -- A common use of the unity archive is for AssetBundles but it can also be used for
 -- Player, Content Archive and ContentDirectory builds.
--- name is UNIQUE: analyze only supports a single build, so two archives with the same name would
--- make queries ambiguous. A duplicate is caught in code and reported (see AnalyzeDuplicateException);
--- the constraint is the durable backstop for that invariant.
+-- name is UNIQUE (case-sensitive, matching the name on the file system): analyze only supports a
+-- single build, so two archives with the same name would make queries ambiguous. A duplicate is
+-- caught in code and reported (see AnalyzeDuplicateException); the constraint is the durable
+-- backstop for that invariant.
 CREATE TABLE IF NOT EXISTS archives
 (
     id INTEGER,

Analyzer/SQLite/Writers/SerializedFileSQLiteWriter.cs

@@ -22,7 +22,9 @@ public class SerializedFileSQLiteWriter : IDisposable
     // Names/ids already written to the archives and serialized_files tables, used to reject a
     // second copy of the same content with a clear error instead of a raw UNIQUE constraint
     // failure. Only a single build can be analyzed at a time (see AnalyzeDuplicateException).
-    private HashSet<string> m_WrittenArchiveNames = new(StringComparer.OrdinalIgnoreCase);
+    // Archive names are compared case-sensitively, matching the archives.name schema constraint
+    // and the name as it exists on the file system.
+    private HashSet<string> m_WrittenArchiveNames = new();
     private HashSet<int> m_WrittenSerializedFileIds = new();
 
     private bool m_SkipReferences;

Documentation/command-analyze.md

@@ -183,7 +183,7 @@ This is expected when the input contains more than one build, and in these commo
 | Player scenes with the same file name (`level0`, …) from different builds | Analyze each build separately |
 
 To compare two builds, analyze each into a separate database and query across them — see
-[Comparing Builds](../../Documentation/comparing-builds.md).
+[Comparing Builds](comparing-builds.md).
 
 ### Slow Analyze times, large output database
 

UnityDataTool.Tests/AnalyzeDuplicateNameTests.cs

@@ -55,8 +55,8 @@ public void Teardown()
     }
 
     // Case 2 / Case 3: two build folders each contain an archive named "assetbundle" (and "scenes").
-    // The second archive of each name is rejected before it is opened, so exactly one row per name
-    // survives and no UNIQUE constraint error is shown.
+    // The second archive of each name is rejected before it is recorded (its contents are never
+    // processed), so exactly one row per name survives and no UNIQUE constraint error is shown.
     [Test]
     public async Task Analyze_ArchivesWithSameName_SkippedWithClearMessage()
     {