aboutsummaryrefslogtreecommitdiff
path: root/text.c
AgeCommit message (Collapse)AuthorFilesLines
2020-10-10text: store blocks in arrayMarc André Tanner1-31/+35
Make block manipulation routines independent of core text data structure, enabling re-usage in different implementations.
2020-10-10text: simplify reading of initial file contentMarc André Tanner1-6/+7
Avoid unnecessary copy and system calls in block_read.
2020-08-29text: provide save function taking a directory descriptorMarc André Tanner1-10/+39
The standard does not specify mkstempat(3). We currently implement it in a non thread safe manner, by temporarily changing the process working directory before invoking mkstemp(3).
2020-08-29text: provide load function taking a directory descriptorMarc André Tanner1-1/+9
2020-06-28text: simplify remapping of original file contentMarc André Tanner1-9/+1
Use mmap with MAP_FIXED which replaces existing mappings without any race conditions between the munmap/mmap calls.
2020-06-22text: remove dead storeMarc André Tanner1-1/+0
2020-06-22text: code cleanup, use local variableMarc André Tanner1-8/+9
No functionl change.
2020-06-22text: fix typo in comments, no code changeMarc André Tanner1-4/+4
2020-05-13text: introduce text_save_method, remove text_save_rangeMarc André Tanner1-9/+9
This utility function is analogous to text_load_method and allows the caller to specify how the file should be saved. It is implemented as a wrapper around the lower level text_save_{begin,write,commit} primitives. The unused text_save_range function has been removed. If needed, use the aforementioned lower level functionality.
2020-01-26text: ignore fsync(2) errors on unsupported directory descriptorsMarc André Tanner1-1/+1
When saving a file by atomically renaming it to its final destination, we fsync(2) the parent directory to make sure the new directory entry is persisted. However, not all file systems support fsync on file descriptors referring to directories. As a result the save operation fails and subsequent attempts result in warnings regarding outdated file content, even though the data has most likely been successfully written. Ignoring this particular error seems fine, because it is a permanent limitation of the file system and not a temporary failure. Fixes #792
2018-05-30text: only default to mmap for files larger than 64 MiBMarc André Tanner1-1/+1
2018-05-30text: allow to specify how the file content should be loadedMarc André Tanner1-3/+5
2018-05-16text: use mkstemp(3) for temporary file creation in atomic savesMarc André Tanner1-6/+25
Instead of simply appending a tilde to the original file name, we now create an unique temporary file based on the pattern `.filename.vis.XXXXXX`. In case the file does not yet exist, we use 0666 & ~umask as permission, (this should match the previous `open(2)` based behavior).
2018-05-16text: do not unlink existing `file~` when saving to `file`Marc André Tanner1-1/+3
Fail atomic save if temporary file already exists. A follow up commit will use `mkstemp(3)` for temporary file creation.
2017-07-05text: limit write(2) calls to INT_MAX bytesMarc André Tanner1-1/+1
Otherwise this fails on macOS. Fix #578
2017-05-03text: remove text_history_get functionMarc André Tanner1-12/+0
As currently implemented this does not properly integrate with multiple cursor support. The functionality should be provided in a layer higher up. The jumplist and changelist need to be redesigned, for now they are broken.
2017-05-03text: remove text_iterate macroMarc André Tanner1-3/+9
2017-05-03text: remove text_insert_newline functionMarc André Tanner1-4/+0
This is no longer needed because we always insert \n never \r\n.
2017-05-03text: remove count argument from text_{earlier,later}Marc André Tanner1-10/+4
2017-05-03text: rename text_sigbus to text_mmapedMarc André Tanner1-2/+3
Add casts to uintptr_t to avoid unrelated pointer comparisons.
2017-05-03text: rename text_byte_get argument nameMarc André Tanner1-2/+2
2017-04-11text: do not redefine _GNU_SOURCEMarc André Tanner1-1/+3
Fix #536
2017-04-09text: use MB_LEN_MAX instead of MB_CUR_MAXMarc André Tanner1-2/+3
2017-04-09vis: remove handling of \r\n line endingsMarc André Tanner1-31/+1
Use something like dos2unix(1) and unix2dos(1), if you need to edit such files.
2017-04-09text: drop special handling of \r\n line endingsMarc André Tanner1-39/+6
2017-04-09text: add mem{r,}chr(3) based byte search functionsMarc André Tanner1-0/+29
These are generally implemented efficiently in libc. While memrchr(3) is non-standard, it is a common extension. If it is not available, we use a simple C implementation from musl.
2017-04-08text: adjust iterator position when switching to adjacent piecesMarc André Tanner1-2/+4
2017-04-08text: introduce text_char_get which converts \r\n to \nMarc André Tanner1-0/+5
2017-04-08text: fix iterator semantics regarding windows style newlinesMarc André Tanner1-2/+4
At some point we might drop this mess and ask users to rely upon dos2unix(1) and unix2dos(1), respectively.
2017-03-31text: Add missing va_end() whenRaúl Peñacoba1-1/+3
vsnprintf fails
2017-03-22text: use proper open(2) flags when saving inplaceMarc André Tanner1-1/+1
This was wrongly changed in commit 74085e92c095d0bf4b98e262cc07ccf9b7dfff3b.
2017-02-27Honor user's umask when saving filesMichael Forney1-2/+2
If the user wants new files to be created as 600, they will have set their umask to 077 in their environment. vis shouldn't restrict this further than the user has configured.
2017-02-27Remove unnecessary umask change in text_save_begin_inplaceMichael Forney1-2/+0
POSIX says that The mkstemp() function shall use the resulting pathname to create the file, and obtain a file descriptor for it, as if by a call to: open(pathname, O_RDWR|O_CREAT|O_EXCL, S_IRUSR|S_IWUSR) So this umask change didn't do anything in practice, unless the original umask was more restrictive than 0177.
2017-02-27Fix a couple typosMichael Forney1-1/+1
2017-02-25text: add iterator accessor function which translates \r\n to \nMarc André Tanner1-0/+11
2017-02-23vis: improve handling of \r\n line endingsMarc André Tanner1-6/+21
Change the text_iterator_char_{prev,next} functions to treat them as a single character, meaning cursor motions will skip both bytes at the same time.
2017-02-20text: fix some integer overflow issuesMarc André Tanner1-3/+6
2017-02-09text: fix use of uninitialized variable in error pathMarc André Tanner1-2/+2
2017-02-08text: clean up iterator semanticsMarc André Tanner1-44/+61
We now guarantee the existence of at least one non-sentinel piece at all time and allow iterators to recover from these delimiting pieces.
2017-02-08text: remove special case for marks at start of fileMarc André Tanner1-4/+0
Unlike EOF address 0 needs no special treatment.
2017-01-20text: fix EOF handling for text_iterator_byte_nextMarc André Tanner1-1/+1
Handle zero length pieces gracefully. At some point we should write a comprehensive set of unit tests for the iterator API.
2016-12-23text: make text_vprintf static, it is only used within text.cMarc André Tanner1-13/+13
2016-12-22text: introduce EMARK to denote an invalid markMarc André Tanner1-2/+2
Technically this macro name is in the reserved namespace of errno.h. The same is true for EPOS. Maybe we should rename them at some point, but for now the short names are convenient. Fix #443 Close #444
2016-12-22text: change datatype of Mark to uintptr_tMarc André Tanner1-4/+6
This should avoid undefined pointer comparisons.
2016-12-22text: make default block size overridable via C pre-processorMarc André Tanner1-0/+2
This will be used for unit test purposes to force more allocations.
2016-12-22text: avoid undefined multiple accesses in expressionMarc André Tanner1-1/+2
Work around for a (bogus?) tis-interpreter warning.
2016-12-22text: do not rely on indeterminate value in a comparisonMarc André Tanner1-1/+1
2016-12-14text: expose text save method to calling codeMarc André Tanner1-9/+5
There are cases where it is useful to specify how the file should be saved.
2016-11-10vis-lua: change misnamed attribute values of `file.newlines`Marc André Tanner1-5/+5
Also rename underlying C code.
2016-11-01text: change usage of sizeof idiom for allocation of new objectsMarc André Tanner1-6/+6
The new code is preferable because it works independently of the variable type. Whereas before a change in type, but not within the sizeof expression would cause a wrongly sized allocation.
You can read more about the author.