concise phrase transform #811
Conversation
| @@ -186,6 +186,27 @@ void redactRegexMatches(String source) { | |||
| } | |||
|
|
|||
|
|
|||
| @SneakyThrows | |||
There was a problem hiding this comment.
test by copilot?? typo PHASE instead of PHRASE everywhere :)
| "Prep Customer Meeting,Prep ", | ||
| "Prep: Customer,Prep: ", | ||
| "Out of the Office: Vacation,Out of the Office", | ||
| "Focus Time,'Focus Time,Focus'", |
There was a problem hiding this comment.
so, this is reality of this; bc transform products ALL matches, not just first. In effect, it's more like categorization than redaction.
behavior is needed if you want "Team weekly" to match both "team" and "weekly" cases. that will end up as team,weekly.
| return patterns.stream() | ||
| .map(p -> p.matcher((String) s)) | ||
| .filter(Matcher::matches) | ||
| .map(m -> m.group(1)) //group 1, bc we created caputuring group in regex above |
There was a problem hiding this comment.
| .map(m -> m.group(1)) //group 1, bc we created caputuring group in regex above | |
| .map(m -> m.group(1)) //group 1, bc we created capturing group in regex above |
| "Focus Time Block,'Focus Time,Focus'", | ||
| "Focus: Secret Project,Focus", | ||
| "No Meeting Wednesday,No Meeting", | ||
| " No Meetings,'No Meetings,No Meetings'", // q: why???? |
There was a problem hiding this comment.
Because matches with No Meeting and No Meetings
There was a problem hiding this comment.
🤔 it shouldn't though ... No Meeting token string shouldn't match No Meetings, bc I'm transforming it into a pattern with \b tokens on either side to only match "word boundaries. So it shouldn't match prefixes.
|
|
||
| List<Pattern> patterns = transform.getAllowedPhrases().stream() | ||
| .map(p -> "\\Q" + p + "\\E") // quote it | ||
| .map(p -> "\\b" + p + "[\\\\s:]*\\b") //boundary match, with optional whitespace or colon at end |
There was a problem hiding this comment.
@jlorper this is where should be limiting to word boundaries.
maybe bug is that "[\\\\s:]*" is wrong, so it's matching character 's' specifically, instead of \s for whitespace as I intended ...
There was a problem hiding this comment.
yeah, it was that - fixed now and works as expected.
* wip of concise phrase transform * fix import * cleanup, fix tests * cleaner rules and examples * fix tests * fix tests * drop case-insensitive multi-pattern stuff * fix pattern to allow spaces rather than s char; and tighten capture pattern * fix tests
Features
Change implications