This solution implements a new config variable MAX_ROWS, which
corresponds to the “Maximum allowed rows to render CSV files. (0 for no
limit)” and rewrites the Render function for CSV files in markup module.
Now the render function only reads the file once, having MAX_FILE_SIZE+1
as a reader limit and MAX_ROWS as a row limit. When the file is larger
than MAX_FILE_SIZE or has more rows than MAX_ROWS, it only renders until
the limit, and displays a user-friendly warning informing that the
rendered data is not complete, in the user's language.
---
Previously, when a CSV file was larger than the limit, the render
function lost its function to render the code. There were also multiple
reads to the file, in order to determine its size and render or
pre-render.
The warning: 
(cherry picked from commit f7125ab61aaa02fd4c7ab0062a2dc9a57726e2ec)
This branch continues the work of #23092 and attempts to rid the
codebase of any `nil` contexts when using a `RenderContext`.
Anything that renders markdown or does post processing may call
`markup.sha1CurrentPatternProcessor()`, and this runs
`git.OpenRepository()`, which needs a context. It will panic if the
context is `nil`. This branch attempts to _always_ include a context
when creating a `RenderContext` to prevent future crashes.
Co-authored-by: Kyle D <kdumontnu@gmail.com>
Change all license headers to comply with REUSE specification.
Fix #16132
Co-authored-by: flynnnnnnnnnn <flynnnnnnnnnn@github>
Co-authored-by: John Olheiser <john.olheiser@gmail.com>
Implements request #14320 The rendering of CSV files does match the diff style.
* Moved CSV logic into base package.
* Added method to create a tabular diff.
* Added CSV compare context.
* Added CSV diff template.
* Use new table style in CSV markup.
* Added file size limit for CSV rendering.
* Display CSV parser errors in diff.
* Lazy read single file.
* Lazy read rows for full diff.
* Added unit tests for various CSV changes.
* detect csv delimiter in csv rendering
fixes #7868
* make linter happy
* fix failing testcase & use ints where possible
* expose markup type to template
previously all markup had the .markdown class, which is incorrect,
as it applies markdown CSS & JS logic to CSV rendering
* fix build (missing `make css`)
* ignore quoted csv content for delimiter scoring
also fix html generation
