Opened 9 years ago
Last modified 7 years ago
#669 reopened defect
Handling accented characters with COinS
| Reported by: | stakats | Owned by: | stakats |
|---|---|---|---|
| Priority: | major | Milestone: | |
| Component: | ingester | Version: | 1.0 |
| Keywords: | Cc: |
Description
We're currently using escape/unescape, and it looks like we should be using encodeURIComponent/decodeURIComponent in ingester.js.
Attachments (1)
Change History (3)
comment:1 Changed 9 years ago by stakats
- Resolution set to fixed
- Status changed from new to closed
Changed 7 years ago by karnesky
comment:2 Changed 7 years ago by karnesky
- Resolution fixed deleted
- Status changed from closed to reopened
COinS does not mandate a character encoding (and OpenURL actually allows a parameter to set the encoding). There are many applications that emit data that is not UTF-8 (older versions of refbase, the OCLC had done this, and I'm sure there are others). They typically use the (deprecated) escape function. Zotero should be more liberal in what it accepts. This patch will use decodeURIComponent, but will fall back to unescape. It works with the examples given in http://forums.zotero.org/discussion/2400 . The patch does not consider what character set the COinS claims to use (there seems to be no reason to), nor does it consider any other escaping scheme (do we need to worry about any?). Minor whitespace flaws are also fixed.
(In [1518]) Closes #669 by supporting encode/decode of non-ASCII (e.g. accented) characters in COinS tags