These are questionable points and errors found during Japanese translation work of the book. Clear and easy errors are corrected in Japanese edition without author's confirmation. Some of them are suspending and waiting author or editor's comment and/or confirmation.
Negative line number means line count from the bottom.
| Page | Line | Section | Category | Original Text and Description | Action | Official |
|---|---|---|---|---|---|---|
| 5 | 16 | Intro | Typo |
For
example, \n is often a shorthand for the newline character,
which is usually LF (012 octal) but can sometimes be CR (15 octal)
depending on the operating system.
015 is better. | fixed | |
| 5 | -5 | Intro | Typesetting | \x{num}
wrong typeface | fixed | |
| 18 | 6 | Perl | Semantic | The match modes (or rack or), /ismxo are locked in.
locking mode "o" sounds strange. | suspending | |
| 19 | 7 | Perl Table 8 | Typesetting | $'
shold be $` | fixed | |
| 19 | 14 | Perl Table 8 | Typo | If true, \m is assumed for all matches without \s.
/m and /s | fixed | yes |
| 20 | 11 | Perl Table 9 | Typo | IsUppper
IsUpper | fixed | |
| 21 | 21 | Perl Example 4 | Questionable | }{<a href="$1">$1</a>}igox;
It's not error, but "o" is not necessary. | suspending | |
| 23 | 12 | Java Table 12 | Questionable | \Z: End of string but before any final line terminator, in any match mode.
Should be or? What does this any mean? | suspending | |
| 25 | 15 | Java Table 14 | Questionable | {n}+: Match at least n times, and never backtrack.
"at least" seems unnecessary. 2003.12.15
By the way, does the expression {n}+ make any sense? How does it backtrack anyway? The way it can backtrack is only giving up whole match, and it happens only when whole pattern is under the effect of other quantifier - (.{n})?, and that quantifier works even for a possessive pattern - (.{n}+)?. 2003.12.16
| NEW! 2003.12.15 Update! 2003.12.16 suspending | |
| 29 | -8 | Java | Trivial | ...such as \p{Lu}, and not \p{Lowercase_Letter}.
"Lowercase_Letter" is "Ll". | Use \p{Ll}. | |
| 36 | 5 | .NET Table 20 | Typesetting | $'
shold be $` | fixed | |
| 37 | 1-4 | .NET | Questionable | First and second interfaces are completely identical. Second is unnecessary or either of them is incorrect. | suspending | |
| 44 | 15,17 | Python Table 22 | Typo | Two A-z should be A-Z. | fixed | |
| 47 | -12 | Python | Semantic | Return string with alphanumerics backslashed so that string can be matched literally.
Should be non-alphanumerics. | fixed | |
| 52 | -6,-7 | PCRE Table 27 | Typo | Two A-z should be A-Z. | fixed | |
| 55 | 17 | PCRE Table 30 | Questionable | {n}+: Match at least n times, and never backtrack.
Same as p.25 Java Table 14 2003.12.17
| NEW! 2003.12.17 suspending | |
| 58 | 25 | PCRE | Semantic | Entry point PCRE uses for pcre_free() calls.
Should be just free(). 2003.12.17
| NEW! 2003.12.17 fixed | |
| 63 | -6,-7 | PHP Table 32 | Typo | Two A-z should be A-Z. | fixed | |
| 64 | 21 | PHP Table 34 | Typo | Table title: Modes
Other tables are using Modifier (Table 6) or Modifier/sequence (Table 13, 18, 24, 29). Probably all titles should be Modifier/sequence. 2003.12.17
| NEW! 2003.12.17 fixed | |
| 66 | -3 | PHP | Semantic | Each additional element N of
matches is an array containing the Nth capture
group match for each full match.
"for each full match" seems unnecessary. Probably copied from previous sentence and forgot to delete. | removed | |
| 70 | -1 | vi Table 37 | Semantic | dot: Any character except newline (unless /s mode).
Does vi have single-line mode? | suspending | |
| 71 | 4-9 | vi Table 37 | Typo | Six "A-z" should be "A-Z". | fixed | |
| 71 | 22-25 | vi Table 37 | Typo | Four lines \i,\L,\u,\U are duplicated.
The first \i must have been copied from \l line and modified partially. | removed | yes |
| 73 | -5 | vi Table 37 | Semantic |
If addr1 is
supplied, substitution will begin on that line number (or the first
matching line) and continue until either the end of the file or the
line indicated (or matched) by addr2.
This description is not correct at least on vim 6.2 and nvi 1.79. Current line is used when either of addr1 and addr2 is omitted. If only addr1 is supplied without following comma, only that line is processed. | suspending | |
| 76 | 3-7 | JavaScript | Typo | Only this table is using hex number preceded by backslash.
I'd like to remove them if there is no specific reason.
2003.12.17
| NEW! 2003.12.17 suspending | |
| 76 | 7 | JavaScript | Typo | \t: Vertical tab, \x0B.
Should be \v. | fixed | |
| 77 | -3 | JavaScript | Typo? | {n}?
Should be {n,}?. 2004.01.06
| NEW! 2004.01.06 fixed | |
| 78 | 8 | JavaScript | Typesetting | /pattern/img.
incorrect typeface. fixed
| | |
| 83 | 13 | Shell Tools | Semantic | [:prop:]
Not wrong but [[:prop:]] is more symmetrical to the next line. | suspending | |
| 83 | 17 | Shell Tools Table 48 | Semantic | Table 48: Anchors and other zero-width testshell tools
"shell tools" seems unnecessary or something is missing between test and shell. | removed | |
| 84 | -2 | Shell Tools | Trivial |
A supplied regular expression must be
defined within the forward slash delimiters (/...).
Seems /.../ is better. | fixed | |
| 85 | -10 | Shell Tools | Typesetting |
The awk script contained in either instructions
or script should be a series of
/pattern/ {action} pairs.
wrong typeface | fixed |