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

[PATCH v2] http-backend: allow empty CONTENT_LENGTH

From
Max Kirillov <max@max630.net>
Date
Sep 7, 2018, 03:36 UTC
Message-ID
<20180907033607.24604-1-max@max630.net>
In-Reply-To
<20180906193516.28909-1-max@max630.net>

According to RFC3875, empty environment variable is equivalent to unset, and for CONTENT_LENGTH it should mean zero body to read.

However, unset CONTENT_LENGTH is also used for chunked encoding to indicate reading until EOF. At least, the test "large fetch-pack requests can be split across POSTs" from t5551 starts faliing, if unset or empty CONTENT_LENGTH is treated as zero length body. So keep the existing behavior as much as possible.

Add a test for the case.
Reported-By: Jelmer Vernooij <jelmer@jelmer.uk>
Signed-off-by: Max Kirillov <max@max630.net>
---
Added the "reported-by" and explained inline the reason to keep existing behavior
 http-backend.c                         |  2 +-
 t/t5562-http-backend-content-length.sh | 11 +++++++++++
 2 files changed, 12 insertions(+), 1 deletion(-)
diff --git a/http-backend.c b/http-backend.c
index e88d29f62b..a1230d7ead 100644
--- a/http-backend.c
+++ b/http-backend.c
@@ -353,7 +353,7 @@ static ssize_t get_content_length(void)
 	ssize_t val = -1;
 	const char *str = getenv("CONTENT_LENGTH");
 
-	if (str && !git_parse_ssize_t(str, &val))
+	if (str && *str && !git_parse_ssize_t(str, &val))
 		die("failed to parse CONTENT_LENGTH: %s", str);
 	return val;
 }
diff --git a/t/t5562-http-backend-content-length.sh b/t/t5562-http-backend-content-length.sh
index 057dcb85d6..ca34c2f054 100755
--- a/t/t5562-http-backend-content-length.sh
+++ b/t/t5562-http-backend-content-length.sh
@@ -152,4 +152,15 @@ test_expect_success 'CONTENT_LENGTH overflow ssite_t' '
 	grep "fatal:.*CONTENT_LENGTH" err
 '
 
+test_expect_success 'empty CONTENT_LENGTH' '
+	env \
+		QUERY_STRING=/repo.git/HEAD \
+		PATH_TRANSLATED="$PWD"/.git/HEAD \
+		GIT_HTTP_EXPORT_ALL=TRUE \
+		REQUEST_METHOD=GET \
+		CONTENT_LENGTH="" \
+		git http-backend <empty_body >act.out 2>act.err &&
+	verify_http_result "200 OK"
+'
+
 test_done
-- 
2.17.0.1185.g782057d875
Previous: Jonathan NiederNext: Jonathan Nieder
Message 12 of 39 in “Re: CONTENT_LENGTH can no longer be empty”
  1. Jonathan NiederSep 6, 2018
  2. http-backend: allow empty CONTENT_LENGTHMax Kirillov, Sep 6, 2018
  3. Junio C HamanoSep 6, 2018
  4. Max KirillovSep 7, 2018
  5. Jeff KingSep 7, 2018
  6. Max KirillovSep 7, 2018
  7. Max KirillovSep 7, 2018
  8. Junio C HamanoSep 7, 2018
  9. Max KirillovSep 8, 2018
  10. Max KirillovSep 9, 2018
  11. Jonathan NiederSep 6, 2018
  12. http-backend: allow empty CONTENT_LENGTHMax Kirillov, Sep 7, 2018
  13. Jonathan NiederSep 8, 2018
  14. http-backend: allow empty CONTENT_LENGTHMax Kirillov, Sep 8, 2018
  15. Jonathan NiederSep 10, 2018
  16. Max KirillovSep 10, 2018
  17. Jonathan NiederSep 11, 2018
  18. http-backend test: make empty CONTENT_LENGTH test more realisticMax Kirillov, Sep 11, 2018
  19. http-backend: allow empty CONTENT_LENGTHMax Kirillov, Sep 8, 2018
  20. http-backend: allow empty CONTENT_LENGTHMax Kirillov, Sep 9, 2018
  21. Jonathan NiederSep 10, 2018
  22. Jeff KingSep 10, 2018
  23. Junio C HamanoSep 10, 2018
  24. Jeff KingSep 10, 2018
  25. http-backend: Treat empty CONTENT_LENGTH as zeroMax Kirillov, Sep 10, 2018
  26. Jonathan NiederSep 10, 2018
  27. Jeff KingSep 11, 2018
  28. Jonathan NiederSep 11, 2018
  29. Jeff KingSep 11, 2018
  30. Jeff KingSep 11, 2018
  31. http-backend: treat empty CONTENT_LENGTH as zeroJonathan Nieder, Sep 11, 2018
  32. Jonathan NiederSep 11, 2018
  33. Junio C HamanoSep 11, 2018
  34. Junio C HamanoSep 11, 2018
  35. Jeff KingSep 12, 2018
  36. Jonathan NiederSep 12, 2018
  37. Junio C HamanoSep 12, 2018
  38. Junio C HamanoSep 11, 2018
  39. Jonathan NiederSep 11, 2018

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.