CLDR-15381 Java formatting with prettier, tabWidth 4, printWidth 100 #1913
Conversation
-Add package.json with custom settings for .java files -Default for printWidth is 80; setting it to 100 results in fewer line breaks -Only java files in tools/cldr-code/src/main/java/org/unicode/cldr/util
|
@macchiati @srl295 this is the same as #1911 except that some lines tend to be longer |
| @@ -216,7 +242,10 @@ public static final class AnnotationSet { | |||
| static final Factory factory = CONFIG.getCldrFactory(); | |||
| static final CLDRFile ENGLISH = CONFIG.getEnglish(); | |||
| static final CLDRFile ENGLISH_ANNOTATIONS = null; | |||
| static final SubdivisionNames englishSubdivisionIdToName = new SubdivisionNames("en", "main"); | |||
There was a problem hiding this comment.
This is definitely splitting lines that it doesn't need to. Maybe set the width higher?
There was a problem hiding this comment.
I'll make another PR with printWidth 120
There was a problem hiding this comment.
See #1919 -- it leaves this line unchanged
| sourceLocale.equals(XMLSource.ROOT_ID) | ||
| ) { | ||
| if ( | ||
| !xpath.equals( |
There was a problem hiding this comment.
In general, I find it harder to read, because it expands to too many lines.
There was a problem hiding this comment.
one way to address that, in this case, might be to assign the string literal "//ldml/characterLabels/characterLabelPattern[@type="category-list"]" to a constant first
There was a problem hiding this comment.
again, see #1919 which leaves this line unchanged
| } | ||
| String flagName = flagLabel == null ? subdivisionName : initialPattern.format(flagLabel, subdivisionName); |
| + "\t" + key | ||
| + "\t" + map.get(key).getShortName() | ||
| + "\t" + Joiner.on(" | ").join(map.get(key).getKeywords())); | ||
| System.out.println( |
There was a problem hiding this comment.
This is really uglier and takes too many vertical lines. Original line breaks were more principled; new ones erase that.
| "👨⚖", "👨🏿⚖", "👩⚖", "👩🏼⚖", | ||
| "👮", "👮♂️", "👮🏼♂️", "👮♀️", "👮🏿♀️", | ||
| "🚴", "🚴🏿", "🚴♂️", "🚴🏿♂️", "🚴♀️", "🚴🏿♀️")) { | ||
| "💏", |
There was a problem hiding this comment.
This one really breaks the intent, since it flattens out the logical structure.
There was a problem hiding this comment.
with this kind of formatter, that's bound to happen -- one approach would be to make the structure more explicit by defining separate arrays and then combining them, and that might be beneficial anyway in terms of separating the code from the data
if we really want a formatter to leave such code unchanged, it will have to be a very different kind of formatter, one that only enforces certain rules and otherwise doesn't change line-breaking or anything else. I don't know if such formatters exist...
| DtdType.platform), keyboards_osx(DtdType.keyboard, | ||
| DtdType.platform), keyboards_und(DtdType.keyboard, DtdType.platform), keyboards_windows(DtdType.keyboard, DtdType.platform), | ||
| ; | ||
| common_annotations(DtdType.ldml), |
There was a problem hiding this comment.
This is much better. The jumbled structure was from an earlier application of a formatter that arbitrarily joined lines.
There was a problem hiding this comment.
this illustrates why we might need to compromise; the formatter has to be allowed to change line-breaking in order to fix this
maybe there are formatters that will break lines but won't join them, I don't know
| // ffi | ||
| { "ffi", "FFI", "FFi", "FfI", "Ffi", "F\uFB01", "fFI", "fFi", "ffI", "ffi", "f\uFB01", "\uFB00I", "\uFB00i", }, | ||
| { | ||
| "ffi", |
There was a problem hiding this comment.
Another bad case. No advantage at all in splitting all these up; just wastes vertical lines.
| R4<K3, K2, K1, V> item = (R4<K3, K2, K1, V>) Row.of(entry0.getKey(), entry1.getKey(), entry2.getKey(), entry2.getValue()); | ||
| for (Entry<Object, Object> entry1 : ( | ||
| (Map<Object, Object>) entry0.getValue() | ||
| ).entrySet()) { |
There was a problem hiding this comment.
Why on earth is it breaking (...).method() before ")."?
There was a problem hiding this comment.
looks like it's balancing parentheses
-Add package.json with custom settings for .java files
-Default for printWidth is 80; setting it to 100 results in fewer line breaks
-Only java files in tools/cldr-code/src/main/java/org/unicode/cldr/util
CLDR-15381