git/list[1] front-page[2] threads[3] people[4] search[5] about
 

[PATCH 12/15] chunk-format: create write_chunks()

From
Derrick Stolee via GitGitGadget <gitgitgadget@gmail.com>
Date
Dec 3, 2020, 16:16 UTC
Message-ID
<03f3255c8f4a953065b2ff8e61816f83534c23ed.1607012215.git.gitgitgadget@gmail.com>
In-Reply-To
<pull.804.git.1607012215.gitgitgadget@gmail.com>
From: Derrick Stolee <dstolee@microsoft.com>

The commit-graph and multi-pack-index files both use a chunk-based file format. They have already unified on using write_table_of_contents(), but we expand upon that by unifying their chunk writing loop.

This takes the concepts already present in the commit-graph that were dropped in the multi-pack-index code during refactoring, including:

* Check the hashfile for how much data was written by each write_fn.
* Allow write_fn() to report an error that results in a failure
  without using die() in the low-level commands.

This simplifies the code in commit-graph.c and midx.c while laying the foundation for future formats using similar ideas.

Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
---
 chunk-format.c | 23 +++++++++++++++++++++++
 chunk-format.h | 13 +++++++++++++
 commit-graph.c | 13 ++-----------
 midx.c         |  3 +--
 4 files changed, 39 insertions(+), 13 deletions(-)
diff --git a/chunk-format.c b/chunk-format.c
index 771b6d98d0..a6643a4fc8 100644
--- a/chunk-format.c
+++ b/chunk-format.c
@@ -24,3 +24,26 @@ void write_table_of_contents(struct hashfile *f,
 	hashwrite_be32(f, 0);
 	hashwrite_be64(f, cur_offset);
 }
+
+int write_chunks(struct hashfile *f,
+		 struct chunk_info *chunks,
+		 int nr,
+		 void *data)
+{
+	int i;
+
+	for (i = 0; i < nr; i++) {
+		uint64_t start_offset = f->total + f->offset;
+		int result = chunks[i].write_fn(f, data);
+
+		if (result)
+			return result;
+
+		if (f->total + f->offset != start_offset + chunks[i].size)
+			BUG("expected to write %"PRId64" bytes to chunk %"PRIx32", but wrote %"PRId64" instead",
+			    chunks[i].size, chunks[i].id,
+			    f->total + f->offset - start_offset);
+	}
+
+	return 0;
+}
diff --git a/chunk-format.h b/chunk-format.h
index 4b9cbeb372..a2c7ddb23b 100644
--- a/chunk-format.h
+++ b/chunk-format.h
@@ -33,4 +33,17 @@ void write_table_of_contents(struct hashfile *f,
 			     struct chunk_info *chunks,
 			     int nr);
 
+/*
+ * Write the data for the given chunk list using the provided
+ * write_fn values. The given 'data' parameter is passed to those
+ * methods.
+ *
+ * The data that is written by each write_fn is checked to be of
+ * the expected size, and a BUG() is thrown if not specified correctly.
+ */
+int write_chunks(struct hashfile *f,
+		 struct chunk_info *chunks,
+		 int nr,
+		 void *data);
+
 #endif
diff --git a/commit-graph.c b/commit-graph.c
index 5494fda1d3..10dcef9d6b 100644
--- a/commit-graph.c
+++ b/commit-graph.c
@@ -1809,17 +1809,8 @@ static int write_commit_graph_file(struct write_commit_graph_context *ctx)
 			num_chunks * ctx->commits.nr);
 	}
 
-	for (i = 0; i < num_chunks; i++) {
-		uint64_t start_offset = f->total + f->offset;
-
-		if (chunks[i].write_fn(f, ctx))
-			return -1;
-
-		if (f->total + f->offset != start_offset + chunks[i].size)
-			BUG("expected to write %"PRId64" bytes to chunk %"PRIx32", but wrote %"PRId64" instead",
-			    chunks[i].size, chunks[i].id,
-			    f->total + f->offset - start_offset);
-	}
+	if (write_chunks(f, chunks, num_chunks, ctx))
+		return -1;
 
 	stop_progress(&ctx->progress);
 	strbuf_release(&progress_title);
diff --git a/midx.c b/midx.c
index 47f5f60fcd..67ac232a81 100644
--- a/midx.c
+++ b/midx.c
@@ -957,8 +957,7 @@ static int write_midx_internal(const char *object_dir, struct multi_pack_index *
 
 	write_table_of_contents(f, header_size, chunks, num_chunks);
 
-	for (i = 0; i < num_chunks; i++)
-		chunks[i].write_fn(f, &ctx);
+	result = write_chunks(f, chunks, num_chunks, &ctx);
 
 	finalize_hashfile(f, NULL, CSUM_FSYNC | CSUM_HASH_IN_STREAM);
 	commit_lock_file(&lk);
-- 
gitgitgadget
Previous: Taylor BlauNext: Taylor Blau
Message 20 of 40 in “Refactor chunk-format into an API”
  1. 00/15 Refactor chunk-format into an APIDerrick Stolee via GitGitGadget, Dec 3, 2020
  2. 02/15 chunk-format: add API for writing table of contentsDerrick Stolee via GitGitGadget, Dec 3, 2020
  3. Taylor BlauDec 8, 2020
  4. 01/15 commit-graph: anonymize data in chunk_write_fnDerrick Stolee via GitGitGadget, Dec 3, 2020
  5. 04/15 midx: use context in write_midx_pack_names()Derrick Stolee via GitGitGadget, Dec 3, 2020
  6. 03/15 midx: rename pack_info to write_midx_contextDerrick Stolee via GitGitGadget, Dec 3, 2020
  7. 05/15 midx: add entries to write_midx_contextDerrick Stolee via GitGitGadget, Dec 3, 2020
  8. Junio C HamanoDec 3, 2020
  9. Derrick StoleeDec 4, 2020
  10. Taylor BlauDec 8, 2020
  11. 07/15 midx: add num_large_offsets to write_midx_contextDerrick Stolee via GitGitGadget, Dec 3, 2020
  12. 06/15 midx: add pack_perm to write_midx_contextDerrick Stolee via GitGitGadget, Dec 3, 2020
  13. 09/15 midx: drop chunk progress during writeDerrick Stolee via GitGitGadget, Dec 3, 2020
  14. 08/15 midx: convert chunk write methods to return intDerrick Stolee via GitGitGadget, Dec 3, 2020
  15. Junio C HamanoDec 3, 2020
  16. Derrick StoleeDec 4, 2020
  17. 11/15 midx: use 64-bit multiplication for chunk sizesDerrick Stolee via GitGitGadget, Dec 3, 2020
  18. Junio C HamanoDec 3, 2020
  19. Taylor BlauDec 8, 2020
  20. 12/15 chunk-format: create write_chunks()Derrick Stolee via GitGitGadget, Dec 3, 2020
  21. Taylor BlauDec 8, 2020
  22. 13/15 chunk-format: create chunk reading APIDerrick Stolee via GitGitGadget, Dec 3, 2020
  23. Junio C HamanoDec 3, 2020
  24. Derrick StoleeDec 4, 2020
  25. Junio C HamanoDec 4, 2020
  26. Junio C HamanoDec 3, 2020
  27. Derrick StoleeDec 4, 2020
  28. 15/15 chunk-format: add technical docsDerrick Stolee via GitGitGadget, Dec 3, 2020
  29. 14/15 commit-graph: restore duplicate chunk checksDerrick Stolee via GitGitGadget, Dec 3, 2020
  30. Derrick StoleeDec 7, 2020
  31. 10/15 midx: use chunk-format API in write_midx_internal()Derrick Stolee via GitGitGadget, Dec 3, 2020
  32. Taylor BlauDec 8, 2020
  33. Derrick StoleeDec 10, 2020
  34. René ScharfeDec 4, 2020
  35. Derrick StoleeDec 4, 2020
  36. Junio C HamanoDec 4, 2020
  37. Taylor BlauDec 8, 2020
  38. René ScharfeDec 9, 2020
  39. Taylor BlauDec 10, 2020
  40. Derrick StoleeDec 10, 2020

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.