edf2csv 0.6.30 → 0.6.32

package/docs/CHANGELOG.md

@@ -3,6 +3,31 @@
 Notable changes to edf2csv. Versions follow [semantic versioning](https://semver.org); while the
 major version is 0, a minor bump may contain breaking changes.
 
+## 0.6.32
+
+### Fixed: the canonical exit-code table had the gap the other three pages had
+
+0.6.10, 0.6.16 and 0.6.18 fixed this on the README, getting-started and the FAQ. This is the
+table those pages defer to, and it had the same two omissions: `--strict` returns 1 over a
+conversion that wrote every file it meant to, and Ctrl-C returns 130.
+
+130 is the sharper miss, because the prose directly under the table already discusses it — "a
+signalled child exits 130 or 143" — while explaining how a batch combines codes. The number was
+on the page, in a paragraph about workers, and absent from the list of what the tool returns.
+
+## 0.6.31
+
+### Fixed: `ANNOTATION_DECODE_FAILED` also covers a duration that decoded fine
+
+A duration of `-3` parses without trouble; what is wrong with it is arithmetic, not decoding. It
+is kept and written to `annotations.csv` exactly as the file gave it — inventing a zero would put
+a number there no writer wrote — and it is warned about, because the recipe this documentation
+gives for the samples an event covers is `onset_s + duration_s`, which for a negative duration
+ends the window before the event starts and selects nothing at all.
+
+The summary line said "couldn't be read", which is true of the other four causes and not of this
+one. Fifth raise site, and the only one where the value survives.
+
 ## 0.6.30
 
 ### Fixed: `DISCONTINUOUS` has six raise sites and the table summarised two

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "edf2csv",
-  "version": "0.6.30",
+  "version": "0.6.32",
   "description": "Convert EDF, EDF+ and BDF biosignal recordings (European Data Format) to CSV from the command line. Local, streaming, and never resamples or alters units.",
   "keywords": [
     "edf",