Commit Graph

7 Commits

Author SHA1 Message Date
oleibman c936f1d9f8
Coverage Improvements (#2859)
Mostly new tests, some code annotations, some minor code changes:
- RichText clone logic is wrong
- TextElement doesn't have object properties, doesn't need clone
2022-06-01 08:29:56 -07:00
Mark Baker 05466e99ce
Html import dimension conversions (#2152)
Allows basic column width conversion when importing from Html that includes UoM... while not overly-sophisticated in converting units to MS Excel's column width units, it should allow import without errors

Also provides a general conversion helper class, and allows column width getters/setters to specify a UoM for easier usage
2021-06-11 17:29:49 +02:00
Adrien Crivelli 49f87de165
Reduce PHPStan error in tests 2021-04-12 11:10:23 +09:00
Adrien Crivelli 6a41381c1d
PSR12 code style 2020-07-26 14:13:11 +09:00
Adrien Crivelli 4739f8b2e7
Merge branch 'readhtml' 2020-07-26 13:11:15 +09:00
Owen Leibman 752a0a5a6c Scrutinizer Recommendations
Two unneeded assignments in tests, one unused parameter in source code.
2020-06-25 23:11:30 -07:00
Owen Leibman 6080c4561d Improve Coverage for HTML Reader
Reader/Html is now covered except for 1 statement.
There is some coverage of RichText when you know in advance that the
html will expand into a single cell.
It is a tougher nut, one that I have not yet cracked,
to try to handle rich text while converting unkown html to multiple cells.
The original author left this as a TODO, and so for now must I.

It made sense to restructure some of the code. There are some changes.
- Issue #1532 is fixed (links are now saved when using rowspan).
- Colors can now be specified as html color name. To accomplish this,
  Helper/Html function colourNameLookup was changed from protected
  to public, and changed to static.
- Superfluous empty lines were eliminated in a number of places, e.g.
  <ul><li>A</li><li>B</li><li>C</li></ul>
  had formerly caused a wrapped cell to be created with 2 empty lines
  followed by A, B, and C on separate lines; it will now just have the
  3 A/B/C lines, which seems like a more sensible interpretation.
- Img alt tag, which had been cast to float, is now used as a string.

Private member "encoding" is not used. Functions getEncoding and setEncoding
have therefore been marked deprecated. In fact, I was unable to get
SecurityScanner to pass *any* html which is not UTF-8. There are
possibly ways of getting around this (in Reader/Html - I have no
intention of messing with Security Scanner), as can be seen in my
companion pull request for Excel2003 Xml Reader. Doing this would be
easier for ASCII-compatible character sets (like ISO-8859-1),
than for non-compatible charsets (like UTF-16). I am not
convinced that the effort is worth it, but am willing to investigate
further.

I added a number of tests, creating an Html directory, and moving
HtmlTest to that directory.
2020-06-25 22:42:38 -07:00