edf2csv 0.6.23 → 0.6.25

package/docs/CHANGELOG.md

@@ -3,6 +3,30 @@
 Notable changes to edf2csv. Versions follow [semantic versioning](https://semver.org); while the
 major version is 0, a minor bump may contain breaking changes.
 
+## 0.6.25
+
+### Fixed: `INPUT_CHANGED` was one row of a summary table and nothing else
+
+Every other diagnostic on the warnings page gets a section: what causes it, what edf2csv does,
+what to do. `INPUT_CHANGED` had a single line in the table at the top and no section anywhere —
+the only code of the thirty-three in that table without one. The docs test that keeps codes and
+pages in step asks whether a code is *named* on the page, which one table row satisfies.
+
+It is not a minor one to leave unexplained: it is the warning that says the output describes the
+file as it was rather than as it is, and the one that turns `--checksum` into a `null`.
+
+## 0.6.24
+
+### Fixed: the timekeeping TAL described as one "which carries an onset and no text"
+
+It is allowed to carry text, writers use that, and edf2csv exports those events like any other —
+which is the whole reason `malformedTimekeepingWithText` is counted apart from
+`malformedTimekeeping`. A 0.5 release exists because the warning beside them said "No event was
+lost" over a conversion whose annotations.csv had gone from six rows to two, and the code comment
+recording that says in as many words that the specification allows both in one entry.
+
+The page describing annotations was the one still asserting they cannot coexist.
+
 ## 0.6.23
 
 ### Fixed: "Record `n` starts at `n * record_duration` seconds, and always will"

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "edf2csv",
-  "version": "0.6.23",
+  "version": "0.6.25",
   "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",