mirror of
https://github.com/storytold/lightcraft.git
synced 2026-11-13 09:20:07 -06:00
1b6e722292
* LR-IMP-ADD-DIALOG: a folder scan counts the folders it could not read and the files it skipped
The bounded folder walk used to drop a folder it could not list, and every file its
caller refused, without a trace. It now counts both (Walked::unreadable and
Walked::skipped); which files are returned, their order, the limits and the
handling of links are unchanged, and the Limits struct is untouched.
import::expand_report keeps those counts with the walk's truncation flag
(Expansion); expand_within delegates to it. ScanOutput carries truncated,
unreadable and unsupported, filled by scan_with, and scan_report gives the
session-backed scan the same output while storing the probes for the import
that follows exactly as scan does.
library.importPreview returns truncated and skipped {unreadable, unsupported}
next to the keys it already had, and says so in its description. A walk that
stops at a bound is still logged once, in expand_report, as before.
* LR-IMP-ADD-DIALOG: tests for the scan's skipped, unreadable and truncated counts
The walk counts files its caller refuses (and not hidden files or the skipped
folder) and a folder that vanishes before it is entered. expand_report, scan_with
and library.importPreview report unsupported files, a walk cut at the depth
limit and an unlistable folder, keep the probes for the import, and leave a
file named on its own uncounted. The unlistable folder uses no permissions on
Unix and a deny entry on Windows, and does nothing where the process can list it
anyway.
* LR-IMP-ADD-DIALOG: the import review says when its scan left something out
When the folder scan stopped at the walk limit, could not read some folders, or
skipped files that are not a supported photo type, the review now shows one line
under the count (for example "Stopped at the scan limit: there may be more
photos · 2 folders could not be read · 3 files skipped (not a supported photo
type)"). It shows nothing when the scan left nothing out. The counts come from
the scan's output and are kept on the dialog.
The two messages with a count go in every formats catalog, as the build requires,
and the plain message in every plain catalog.
* LR-IMP-ADD-DIALOG: tests for the import review's scan note
Opening the review for a folder with a stray text file, for a tree deeper than
the walk goes, and for a folder of photos only shows the matching line, or none.
The line follows the counters while the review is open, is written in the
language in use, and names singular and plural counts correctly.
* LR-IMP-ADD-DIALOG: tracker note for the import review's scan note
The row says what the review reports when its scan left something out, what
library.importPreview returns for it, and which other folder walks do not report
these counts. Revision history and last-updated line added; the status stays
done and the generated summary is not touched.
* LR-IMP-ADD-DIALOG: XMP sidecars are not counted as unsupported files
The import reads a photo's XMP sidecar along with the photo, so a sidecar is not a
file the scan skipped. expand_report now leaves .xmp files (any letter case) out of
Expansion::unsupported, which feeds ScanOutput, the "skipped.unsupported" count of
library.importPreview and the review's note; every other file that is not a
supported photo type still counts. The walk itself stays generic: Walked::skipped
still counts every file its caller refused, and which files are returned is
unchanged.
* LR-IMP-ADD-DIALOG: tests for leaving XMP sidecars out of the unsupported count
A folder with a photo, sidecars in lower, upper and mixed case (one in a subfolder)
and a text file reports one unsupported file through the expansion, the scan and
library.importPreview; a folder of photos and sidecars only reports none. The review
shows no note for sidecars alone and the text file's line once one is added.
* LR-IMP-ADD-DIALOG: tracker wording for the sidecar exclusion
The row and the revision-history entry say that XMP sidecars are not counted among
the files the review reports as skipped.
* rename importPreview's walk-truncation flag to scanLimited
- avoids one reply using truncated for two meanings once #2705/#2873 land (maintainer request)
* ui: translate the import scan notes into cs, fi, id, it, ka, pl, sv and vi
* fix: register the scan note against the Ui, not the Context (main's register takes &Ui)
* Report diagnostics even when the scan finds no photos
When a scan finds no photos, the message now says what it left out: folders it could not read, files that are not photos, or a walk stopped at its limit. A complete scan with no photos keeps the plain message.
* Format the scan results the merge touched
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: eliets2 <287273517+eliets2@users.noreply.github.com>
Co-authored-by: FredApps <153173854+FredApps@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>