summaryrefslogtreecommitdiff
path: root/language/hyphenation/ruhyphen/BUGS
blob: 84456e41f64801549c397db7139a0441e977a2e8 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
-*-coding: koi8-r;-*-

1) the following command shows useless (because of \lefthyphenmin=2)
or incorrect patterns which set hyphenation after the first letter:

grep '^\..[0-9]' ruhyph*.tex

ruhyphas.tex:
.×8Ó8Ô8Ò8
.Ï5×Ø4
.Ï4Ô×ÅÒÄ
.Ï4Ô×ÅÒÖÄ

2) the following command shows useless (because of \righthyphenmin=2)
or incorrect patterns which set hyphenation before the last letter:

grep '[0-9].\.$' ruhyph*.tex

ruhyphas.tex:
.ÕÌØ5Å.
.Ä×Ï5Å6Ë.

3) the patterns for the whole words shown by the command

grep '^\..*\.$' ruhyph*.tex

are also questionable (probably it is better to store them in
\hyphenation list). Such patterns occur in ruhyphas.tex and
ruhyphzn.tex.

4) the following command will show patterns which contain letter
combinations absent in the full dictionary:

./reduce-patt < ruhyphXX.tex | diff - ruhyphXX.tex | perl -n -e 'print if (s/^> //)'

where XX should be replaced accordingly, and it is assumed that the
script `reduce-patt' refers to the full lowercased wordlist (e.g.
generated from the rus-ispell dictionary by Alexander Lebedev).

There may be 3 reasons for such `wrong' patterns:

  a) there are no existing words which could be influenced by these
     patterns. In this case, such patterns should be removed.

  b) the pattern corresponds to the word with letter `cyre', but
     the orthographically-correct word uses the letter `cyryo'.
     In this case, the pattern should be leaved, because it
     corresponds to the (relatively common) case when `cyryo' is
     replaced by `cyre'.

  c) the patterns are correct (correspond to existing words), but
     the dictionary does not contain such words. In this case, the
     dictionary should be supplemented.

5) to ensure that there will be absolutely no -xx breaks for English
words when the combined Russian-English hyphenation is used, a changed
English patterns should be used, with removed `strange' patterns
matching the regexp "[0-9][a-z][a-z]\." (see the file enrhm2.tex).

6) [not a bug] should the main hyphenation files be sorted using
`sorthyph' script?

7) Known cases of bad hyphenation (to be fixed by authors of
corresponding patterns). All found hyphenation positions are shown.
Bad positions are marked with `='. Please send additional examples of
bad hyphenation for various patterns in this collection to email
addresses given at the end of README. See also some other examples
in the release notes of Alexander Lebedev's patterns in the file
README.ruhyphal.

ruhyphas.tex:

  ÓÉÎ-È-ÒÏÔ=ÒÏÎ-ÎÙÊ
  ÍÅ=ÖÁ=ÔÏÍ-ÎÙÊ
  ÐÏÄ-Ô×ÅÒÖ=ÄÁ-ÀÔ