Many hyperlinks are disabled.
Use anonymous login
to enable hyperlinks.
28 check-ins related to "bomRefactor"
|
2013-03-19
| ||
| 17:40 | Merge UTF-16 byte swapping fix and test-looks-like-utf command enhancements. check-in: b4bec3753d user: mistachkin tags: trunk | |
| 17:37 | Fix overly eager byte swapping when checking for UTF-16 text. Closed-Leaf check-in: 3f2f1e62fa user: mistachkin tags: bomRefactor | |
| 08:34 | Fix expected value of test-cases: The value of LOOK_LONE_CR is wrong in 19 cases. I leave it to Joe to fix the code. check-in: 8af1541ac5 user: jan.nijtmans tags: bomRefactor | |
|
2013-03-18
| ||
| 23:47 | Make sure that LOOK_CR is set even when a CR/LF pair is detected. Rename the LOOK_LENGTH flag to LOOK_LONG for clarify. Add LOOK_SHORT flag to indicate that the looks_like_utf16() function did not perform a full check. Support tests for UTF-16 in reverse byte order. Enhancements to the test-looks-like-utf command. check-in: b0b3f2a44f user: mistachkin tags: bomRefactor | |
| 12:37 | Adapt test-case 112 such that it contains a reversed CR/LF, a case not covered before. Fix detection of reversed CR/LF and lone CR in reversed UTF-16 case, broken by [e3f9a42b58]. check-in: a4cdc7235a user: jan.nijtmans tags: trunk | |
|
2013-03-15
| ||
| 12:29 | merge trunk Closed-Leaf check-in: 59f2644788 user: jan.nijtmans tags: bomRefactor | |
| 12:23 | Don't let looks_like_utf8/16 decide any more whether the blob is text or binary. Calling code can do that based on the returned flags. This simplifies looks_like_utf8/16 a lot. check-in: 276b34955b user: jan.nijtmans tags: trunk | |
| 11:16 | minor simplifications check-in: 731e07dd94 user: jan.nijtmans tags: bomRefactor | |
| 10:28 | fix comments check-in: d0a36540fd user: jan.nijtmans tags: bomRefactor | |
| 10:17 | re-apply [3d6d2e7f35]: Better handling and tests for initial CR/LF chars in the looks_like_utf*() functions. check-in: 3528062ac1 user: jan.nijtmans tags: bomRefactor | |
| 10:05 | Add two new utf-16 test-cases, which were handled wrong up to now, and fix the code accordingly. check-in: fab2b8aea5 user: jan.nijtmans tags: trunk | |
|
2013-03-13
| ||
| 15:03 | merge trunk Closed-Leaf check-in: 9263d4c2c3 user: jan.nijtmans tags: bomRefactor | |
| 14:52 | make expected "utf-check" parameters less cryptic check-in: 5a9f855810 user: jan.nijtmans tags: trunk | |
| 13:44 | remove debug code check-in: 7ef1890c19 user: jan.nijtmans tags: bomRefactor | |
| 13:37 | merge trunk check-in: 3718383e5c user: jan.nijtmans tags: bomRefactor | |
| 01:22 | Update the built-in SQLite to the latest 3.7.16 beta from upstream. check-in: f803d534d0 user: drh tags: trunk | |
|
2013-03-11
| ||
| 15:37 | Combine looks_like_utf8/looks_like_utf16 to a single looks_like_utf function, which can be used in both checkin.c and utf.test. This assures that utf.test really checks the same thing as checkin.c is using. check-in: e42ea98f95 user: jan.nijtmans tags: bomRefactor | |
|
2013-03-10
| ||
| 13:32 | Add possible 'a=all' response to "fossil clean" check-in: 5f4881d421 user: jan.nijtmans tags: trunk | |
|
2013-03-08
| ||
| 11:56 | merge trunk Closed-Leaf check-in: 8be674a133 user: jan.nijtmans tags: bomRefactor | |
| 09:15 | Fix crnl-warning for file which starts with þÿ or ÿþ (latin-1) and has an odd number of bytes. check-in: 25eec8cfaf user: jan.nijtmans tags: trunk | |
|
2013-03-07
| ||
| 22:26 | Detect overlong and nul-bytes for files starting with utf-16 BOM but with odd number of bytes. Closed-Leaf check-in: 14019cb0e2 user: jan.nijtmans tags: bomRefactor | |
| 10:19 | merge-mark trunk check-in: 9baee7e90f user: jan.nijtmans tags: bomRefactor | |
| 10:16 | Fix gcc warning, and add test-case (which fails in current trunk) check-in: c53b128f15 user: jan.nijtmans tags: bomRefactor | |
| 09:56 | Since the blob size is unsigned, looks_like_utf16() cannot easily process the content if the size is not even. check-in: 32b8fc7dcd user: mistachkin tags: trunk | |
| 09:21 | Better comment. check-in: 2237e34377 user: jan.nijtmans tags: bomRefactor | |
| 08:54 | Forgot to remove the odd bytes check, and fixed comment. check-in: b5a3f0b48a user: jan.nijtmans tags: bomRefactor | |
| 08:21 | If the blob has an odd number of bytes, it means that the content of the file cannot be interpreted as UTF-16 characters. So, starts_with_utf16_bom() should return false and looks_like_utf16() shouldn't even be called. The flag LOOK_ODD makes no sense in starts_with_utf16_bom(). check-in: 7b0b7ec87d user: jan.nijtmans tags: bomRefactor | |
| 01:27 | Add support for detecting standalone carriage-return characters via the looks_like_utf*() functions. check-in: 6c7185d9ba user: mistachkin tags: trunk | |