From ababfe80efee2df3e9ec2e1fae5064b5d91c3aff Mon Sep 17 00:00:00 2001
From: Daniel Ehrenberg <littledan@Macintosh-122.local>
Date: Mon, 16 Mar 2009 17:53:38 -0500
Subject: [PATCH 1/8] More regexp docs; unix line ending support
---
basis/regexp/ast/ast.factor | 3 +-
basis/regexp/classes/classes.factor | 2 +-
basis/regexp/compiler/compiler.factor | 9 ++-
basis/regexp/nfa/nfa.factor | 9 ++-
basis/regexp/parser/parser.factor | 36 +++++------
basis/regexp/regexp-docs.factor | 86 ++++++++++++++++++++++++---
basis/regexp/regexp-tests.factor | 10 ++++
7 files changed, 118 insertions(+), 37 deletions(-)
diff --git a/basis/regexp/ast/ast.factor b/basis/regexp/ast/ast.factor
index ffaed2db62..1c11ed5c7d 100644
--- a/basis/regexp/ast/ast.factor
+++ b/basis/regexp/ast/ast.factor
@@ -37,8 +37,7 @@ C: <with-options> with-options
TUPLE: options on off ;
C: <options> options
-SINGLETONS: unix-lines dotall multiline comments case-insensitive
-unicode-case reversed-regexp ;
+SINGLETONS: unix-lines dotall multiline case-insensitive reversed-regexp ;
: <maybe> ( term -- term' )
f <concatenation> 2array <alternation> ;
diff --git a/basis/regexp/classes/classes.factor b/basis/regexp/classes/classes.factor
index d26ff7f69c..e3a1774585 100644
--- a/basis/regexp/classes/classes.factor
+++ b/basis/regexp/classes/classes.factor
@@ -12,7 +12,7 @@ ascii-class punctuation-class java-printable-class blank-class
control-character-class hex-digit-class java-blank-class c-identifier-class
unmatchable-class terminator-class word-boundary-class ;
-SINGLETONS: beginning-of-input ^ end-of-input $ end-of-file word-break ;
+SINGLETONS: beginning-of-input ^ end-of-input $ end-of-file ^unix $unix word-break ;
TUPLE: range from to ;
C: <range> range
diff --git a/basis/regexp/compiler/compiler.factor b/basis/regexp/compiler/compiler.factor
index b55cab6294..95511965d1 100644
--- a/basis/regexp/compiler/compiler.factor
+++ b/basis/regexp/compiler/compiler.factor
@@ -17,9 +17,6 @@ SYMBOL: backwards?
M: t question>quot drop [ 2drop t ] ;
M: f question>quot drop [ 2drop f ] ;
-M: not-class question>quot
- class>> question>quot [ not ] compose ;
-
M: beginning-of-input question>quot
drop [ drop zero? ] ;
@@ -40,6 +37,12 @@ M: $ question>quot
M: ^ question>quot
drop [ { [ drop zero? ] [ [ 1- ] dip ?nth "\r\n" member? ] } 2|| ] ;
+M: $unix question>quot
+ drop [ { [ length = ] [ ?nth CHAR: \n = ] } 2|| ] ;
+
+M: ^unix question>quot
+ drop [ { [ drop zero? ] [ [ 1- ] dip ?nth CHAR: \n = ] } 2|| ] ;
+
M: word-break question>quot
drop [ word-break-at? ] ;
diff --git a/basis/regexp/nfa/nfa.factor b/basis/regexp/nfa/nfa.factor
index 20be6b87d8..d59d4818ec 100644
--- a/basis/regexp/nfa/nfa.factor
+++ b/basis/regexp/nfa/nfa.factor
@@ -60,11 +60,16 @@ GENERIC: modify-epsilon ( tag -- newtag )
M: object modify-epsilon ;
+: line-option ( multiline unix-lines default -- option )
+ multiline option? [
+ drop [ unix-lines option? ] 2dip swap ?
+ ] [ 2nip ] if ;
+
M: $ modify-epsilon
- multiline option? [ drop end-of-input ] unless ;
+ $unix end-of-input line-option ;
M: ^ modify-epsilon
- multiline option? [ drop beginning-of-input ] unless ;
+ ^unix beginning-of-input line-option ;
M: tagged-epsilon nfa-node
clone [ modify-epsilon ] change-tag add-simple-entry ;
diff --git a/basis/regexp/parser/parser.factor b/basis/regexp/parser/parser.factor
index c6a69f2508..7b2d6af2c1 100644
--- a/basis/regexp/parser/parser.factor
+++ b/basis/regexp/parser/parser.factor
@@ -2,7 +2,7 @@
! See http://factorcode.org/license.txt for BSD license.
USING: peg.ebnf kernel math.parser sequences assocs arrays fry math
combinators regexp.classes strings splitting peg locals accessors
-regexp.ast ;
+regexp.ast unicode.case ;
IN: regexp.parser
: allowed-char? ( ch -- ? )
@@ -19,20 +19,19 @@ ERROR: bad-number ;
ERROR: bad-class name ;
: name>class ( name -- class )
- {
- { "Lower" letter-class }
- { "Upper" LETTER-class }
- { "Alpha" Letter-class }
- { "ASCII" ascii-class }
- { "Digit" digit-class }
- { "Alnum" alpha-class }
- { "Punct" punctuation-class }
- { "Graph" java-printable-class }
- { "Print" java-printable-class }
- { "Blank" non-newline-blank-class }
- { "Cntrl" control-character-class }
- { "XDigit" hex-digit-class }
- { "Space" java-blank-class }
+ >string >case-fold {
+ { "lower" letter-class }
+ { "upper" LETTER-class }
+ { "alpha" Letter-class }
+ { "ascii" ascii-class }
+ { "digit" digit-class }
+ { "alnum" alpha-class }
+ { "punct" punctuation-class }
+ { "graph" java-printable-class }
+ { "blank" non-newline-blank-class }
+ { "cntrl" control-character-class }
+ { "xdigit" hex-digit-class }
+ { "space" java-blank-class }
! TODO: unicode-character-class
} [ bad-class ] at-error ;
@@ -66,11 +65,8 @@ ERROR: bad-class name ;
{ CHAR: i case-insensitive }
{ CHAR: d unix-lines }
{ CHAR: m multiline }
- { CHAR: n multiline }
{ CHAR: r reversed-regexp }
{ CHAR: s dotall }
- { CHAR: u unicode-case }
- { CHAR: x comments }
} ;
: ch>option ( ch -- singleton )
@@ -101,8 +97,8 @@ CharacterInBracket = !("}") Character
QuotedCharacter = !("\\E") .
-Escape = "p{" CharacterInBracket*:s "}" => [[ s >string name>class <primitive-class> ]]
- | "P{" CharacterInBracket*:s "}" => [[ s >string name>class <primitive-class> <negation> ]]
+Escape = "p{" CharacterInBracket*:s "}" => [[ s name>class <primitive-class> ]]
+ | "P{" CharacterInBracket*:s "}" => [[ s name>class <primitive-class> <negation> ]]
| "Q" QuotedCharacter*:s "\\E" => [[ s <concatenation> ]]
| "u" Character:a Character:b Character:c Character:d
=> [[ { a b c d } hex> ensure-number ]]
diff --git a/basis/regexp/regexp-docs.factor b/basis/regexp/regexp-docs.factor
index b35f8d1cf3..a7cb0a3715 100644
--- a/basis/regexp/regexp-docs.factor
+++ b/basis/regexp/regexp-docs.factor
@@ -33,20 +33,71 @@ ARTICLE: { "regexp" "construction" } "Constructing regular expressions"
"Another approach is to use " { $vocab-link "regexp.combinators" } "." ;
ARTICLE: { "regexp" "syntax" } "Regular expression syntax"
-"Regexp syntax is largely compatible with Perl, Java and extended POSIX regexps, but not completely. A new addition is the inclusion of a negation operator, with the syntax " { $snippet "(?~foo)" } " to match everything that does not match " { $snippet "foo" } "."
+"Regexp syntax is largely compatible with Perl, Java and extended POSIX regexps, but not completely. Below, the syntax is documented."
{ $heading "Characters" }
+"At its core, regular expressions consist of character literals. For example, " { $snippet "R/ f/" } " is a regular expression matching just the string 'f'. In addition, the normal escape codes are provided, like " { $snippet "\\t" } " for the tab character and " { $snippet "\\uxxxxxx" } "for an arbitrary Unicode code point, by its hex value. In addition, any character can be preceded by a backslash to escape it, unless this has special meaning. For example, to match a literal opening parenthesis, use " { $snippet "\\(" } "."
+{ $heading "Concatenation, alternation and grouping" }
+"Regular expressions can be built out of multiple characters by concatenation. For example, " { $snippet "R/ ab/" } " matches a followed by b. The " { $snippet "|" } " (alternation) operator can construct a regexp which matches one of two alternatives. Parentheses can be used for gropuing. So " { $snippet "R/ f(oo|ar)/" } " would match either 'foo' or 'far'."
{ $heading "Character classes" }
+"Square brackets define a convenient way to refer to a set of characters. For example, " { $snippet "[ab]" } " refers to either a or b. And " { $snippet "[a-z]" } " refers to all of the characters between a and z, in code point order. You can use these together, as in " { $snippet "[ac-fz]" } " which matches all of the characters between c and f, in addition to a and z. Character classes can be negated using a carat, as in " { $snippet "[^a]" } " which matches all characters which are not a."
{ $heading "Predefined character classes" }
+"Several character classes are predefined, both for convenience and because they are too large to represent directly. In Factor regular expressions, all character classes are Unicode-aware."
+{ $table
+ { { $snippet "\\d" } "Digits" }
+ { { $snippet "\\D" } "Not digits" }
+ { { $snippet "\\s" } "Whitespace" }
+ { { $snippet "\\S" } "Not whitespace" }
+ { { $snippet "\\w" } "Word character (alphanumeric or underscore)" }
+ { { $snippet "\\W" } "Not word character" }
+ { { $snippet "\\p{property}" } "Character which fulfils the property" }
+ { { $snippet "\\P{property}" } "Character which does not fulfil the property" } }
+"Properties for " { $snippet "\\p" } " and " { $snippet "\\P" } " (case-insensitive):"
+{ $table
+ { { $snippet "\\p{lower}" } "Lower case letters" }
+ { { $snippet "\\p{upper}" } "Upper case letters" }
+ { { $snippet "\\p{alpha}" } "Letters" }
+ { { $snippet "\\p{ascii}" } "Characters in the ASCII range" }
+ { { $snippet "\\p{alnum}" } "Letters or numbers" }
+ { { $snippet "\\p{punct}" } "Punctuation" }
+ { { $snippet "\\p{blank}" } "Non-newline whitespace" }
+ { { $snippet "\\p{cntrl}" } "Control character" }
+ { { $snippet "\\p{space}" } "Whitespace" }
+ { { $snippet "\\p{xdigit}" } "Hexidecimal digit" } } ! In the future: Unicode
+"Full unicode properties are not yet supported."
{ $heading "Boundaries" }
+"Special operators exist to match certain points in the string. These are called 'zero-width' because they do not consume any characters."
+{ $table
+ { { $snippet "^" } "Beginning of a line" }
+ { { $snippet "$" } "End of a line" }
+ { { $snippet "\\A" } "Beginning of text" }
+ { { $snippet "\\z" } "End of text" }
+ { { $snippet "\\Z" } "Almost end of text: only thing after is newline" }
+ { { $snippet "\\b" } "Word boundary (by Unicode word boundaries)" }
+ { { $snippet "\\b" } "Not word boundary (by Unicode word boundaries)" } }
{ $heading "Greedy quantifiers" }
-{ $heading "Reluctant quantifiers" }
-{ $heading "Posessive quantifiers" }
-{ $heading "Logical operations" }
+"It is possible to have a regular expression which matches a variable number of occurrences of another regular expression."
+{ $table
+ { { $snippet "a*" } "Zero or more occurrences of a" }
+ { { $snippet "a+" } "One or more occurrences of a" }
+ { { $snippet "a?" } "Zero or one occurrences of a" }
+ { { $snippet "a{n}" } "n occurrences of a" }
+ { { $snippet "a{n,}" } "At least n occurrences of a" }
+ { { $snippet "a{,m}" } "At most m occurrences of a" }
+ { { $snippet "a{n,m}" } "Between n and m occurrences of a" } }
+"All of these quantifiers are " { $emphasis "greedy" } ", meaning that they take as many repetitions as possible within the larger regular expression. Reluctant and posessive quantifiers are not yet supported."
{ $heading "Lookaround" }
+"Operators are provided to look ahead and behind the current point in the regular expression. These can be used in any context, but they're the most useful at the beginning or end of a regular expression."
+{ $table
+ { { $snippet "(?=a)" } "Asserts that the current position is immediately followed by a" }
+ { { $snippet "(?!a)" } "Asserts that the current position is not immediately followed by a" }
+ { { $snippet "(?<=a)" } "Asserts that the current position is immediately preceded by a" }
+ { { $snippet "(?<!a)" } "Asserts that the current position is not immediately preceded by a" } }
+{ $heading "Quotation" }
+"To make it convenient to have a long string which uses regexp operators, a special syntax is provided. If a substring begins with " { $snippet "\\Q" } " then everything until " { $snippet "\\E" } " is quoted (escaped). For example, " { $snippet "R/ \\Qfoo\\bar|baz()\\E/" } " matches exactly the string " { $snippet "\"foo\\bar|baz()\"" } "."
{ $heading "Unsupported features" }
"One missing feature is backreferences. This is because of a design decision to allow only regular expressions following the formal theory of regular languages. For more information, see " { $link { "regexp" "theory" } } ". You can create a new regular expression to match a particular string using " { $vocab-link "regexp.combinators" } " and group capture is available to extract parts of a regular expression match." $nl
"Another feature is Perl's " { $snippet "\\G" } " syntax, which references the previous match, is not included. This is because that sequence is inherently stateful, and Factor regexps don't hold state." $nl
-"Additionally, none of the operations which embed code into a regexp are supported, as this would require the inclusion of the Factor parser and compiler in any application which wants to expose regexps to the user. None of the casing operations are included, for simplicity." ; ! Also describe syntax, from the beginning
+"None of the operations which embed code into a regexp are supported, as this would require the inclusion of the Factor parser and compiler in any application which wants to expose regexps to the user. None of the casing operations are included of Perl like \\L, for simplicity." ; ! Also describe syntax, from the beginning
ARTICLE: { "regexp" "options" } "Regular expression options"
"When " { $link { "regexp" "construction" } } ", various options can be provided. Options have single-character names. A string of options has one of the following two forms:"
@@ -58,13 +109,30 @@ $nl
{ "i" { $link case-insensitive } }
{ "d" { $link unix-lines } }
{ "m" { $link multiline } }
- { "n" { $link multiline } }
- { "r" { $link reversed-regexp } }
{ "s" { $link dotall } }
- { "u" { $link unicode-case } }
- { "x" { $link comments } }
+ { "r" { $link reversed-regexp } }
} ;
+HELP: case-insensitive
+{ $syntax "R/ .../i" }
+{ $description "On regexps, the " { $snippet "i" } " option makes the match case-insenstive. Currently, this is handled incorrectly with respect to Unicode, as characters like ß do not expand into SS in upper case. This should be fixed in a future version." } ;
+
+HELP: unix-lines
+{ $syntax "R/ .../d" }
+{ $description "With this mode, only newlines (" { $snippet "\\n" } ") are recognized for line breaking. This affects " { $snippet "$" } " and " { $snippet "^" } " when in multiline mode." } ;
+
+HELP: multiline
+{ $syntax "R/ .../m" }
+{ $description "This mode makes the zero-width constraints " { $snippet "$" } " and " { $snippet "^" } " match the beginning or end of a line. Otherwise, they only match the beginning or end of the input text. This can be used together with " { $link dotall } "." } ;
+
+HELP: dotall
+{ $syntax "R/ .../s" }
+{ $description "This mode, traditionally called single line mode, makes " { $snippet "." } " match everything, including line breaks. By default, it does not match line breaking characters. This can be used together with " { $link multiline } "." } ;
+
+HELP: reversed-regexp
+{ $syntax "R/ .../r" }
+{ $description "When running a regexp compiled with this mode, matches will start from the end of the input string, going towards the beginning." } ;
+
ARTICLE: { "regexp" "theory" } "The theory of regular expressions"
"Far from being just a practical tool invented by Unix hackers, regular expressions were studied formally before computer programs were written to process them." $nl
"A regular language is a set of strings that is matched by a regular expression, which is defined to have characters and the empty string, along with the operations concatenation, disjunction and Kleene star. Another way to define the class of regular languages is as the class of languages which can be recognized with constant space overhead, ie with a DFA. These two definitions are provably equivalent." $nl
diff --git a/basis/regexp/regexp-tests.factor b/basis/regexp/regexp-tests.factor
index a449b3e2f0..0836c0988b 100644
--- a/basis/regexp/regexp-tests.factor
+++ b/basis/regexp/regexp-tests.factor
@@ -470,3 +470,13 @@ IN: regexp-tests
[ t ] [ "abcdefg" "a(?:bcdefg)" <regexp> matches? ] unit-test
[ 3 ] [ "caba" "(?<=b)a" <regexp> first-match from>> ] unit-test
+
+[ t ] [ "\ra" R/ .^a/ms matches? ] unit-test
+[ f ] [ "\ra" R/ .^a/mds matches? ] unit-test
+[ t ] [ "\na" R/ .^a/ms matches? ] unit-test
+[ t ] [ "\na" R/ .^a/mds matches? ] unit-test
+
+[ t ] [ "a\r" R/ a$./ms matches? ] unit-test
+[ f ] [ "a\r" R/ a$./mds matches? ] unit-test
+[ t ] [ "a\n" R/ a$./ms matches? ] unit-test
+[ t ] [ "a\n" R/ a$./mds matches? ] unit-test
From b745930b2875797d24cab2aac71adc57bb57e1f0 Mon Sep 17 00:00:00 2001
From: Daniel Ehrenberg <littledan@Macintosh-122.local>
Date: Mon, 16 Mar 2009 18:14:39 -0500
Subject: [PATCH 2/8] More regexp docs
---
basis/regexp/regexp-docs.factor | 16 +++++++++++++---
1 file changed, 13 insertions(+), 3 deletions(-)
diff --git a/basis/regexp/regexp-docs.factor b/basis/regexp/regexp-docs.factor
index a7cb0a3715..041c78380b 100644
--- a/basis/regexp/regexp-docs.factor
+++ b/basis/regexp/regexp-docs.factor
@@ -1,6 +1,7 @@
! Copyright (C) 2008, 2009 Doug Coleman, Daniel Ehrenberg.
! See http://factorcode.org/license.txt for BSD license.
-USING: kernel strings help.markup help.syntax math regexp.parser regexp.ast ;
+USING: kernel strings help.markup help.syntax math regexp.parser
+regexp.ast multiline ;
IN: regexp
ABOUT: "regexp"
@@ -21,8 +22,17 @@ ARTICLE: "regexp" "Regular expressions"
{ $subsection { "regexp" "deploy" } } ;
ARTICLE: { "regexp" "intro" } "A quick introduction to regular expressions"
-
-;
+"Regular expressions are a terse way to do certain simple string processing tasks. For example, to replace all instances of " { $snippet "foo" } " in one string with { $snippet "bar" } ", the following can be used:
+{ $code "R/ foo/ \"bar\" re-replace" }
+"That could be done with sequence operations, but consider doing this replacement for an arbitrary number of o's, at least two:"
+{ $code "R/ foo+/ \"bar\" re-replace" }
+"The " { $snippet "+" } " operator matches one or more occurrences of the previous expression; in this case " { $snippet "o" } ". Another useful feature is alternation. Say we want to do this replacement with fooooo or boooo. Then we could use the code"
+{ $code "R/ (f|b)oo+/ \"bar\" re-replace" }
+"To search a file for all lines that match a given regular expression, you could use code like this:"
+{ $code <" "file.txt" ascii file-lines [ R/ (f|b)oo+/ re-contains? ] filter "> }
+"To test if a string in its entirity matches a regular expression, the following can be used:"
+{ $example <" "fooo" R/ (b|f)oo+/ matches? . "> "t" }
+"Regular expressions can't be used for all parsing tasks. For example, they are not powerful enough to match balancing parentheses." ;
ARTICLE: { "regexp" "construction" } "Constructing regular expressions"
"Most of the time, regular expressions are literals and the parsing word should be used, to construct them at parse time. This ensures that they are only compiled once, and gives parse time syntax checking."
From 7a0ce748df5cc57ffe2c34014d5be1846c469223 Mon Sep 17 00:00:00 2001
From: Daniel Ehrenberg <littledan@Macintosh-122.local>
Date: Mon, 16 Mar 2009 18:28:15 -0500
Subject: [PATCH 3/8] Cleaning up XML to make : string>xml <string-reader>
read-xml ;
---
basis/xml/tests/test.factor | 1 +
basis/xml/xml.factor | 14 ++++++++++----
2 files changed, 11 insertions(+), 4 deletions(-)
diff --git a/basis/xml/tests/test.factor b/basis/xml/tests/test.factor
index 818a28c892..1d07aa9406 100644
--- a/basis/xml/tests/test.factor
+++ b/basis/xml/tests/test.factor
@@ -74,3 +74,4 @@ SYMBOL: xml-file
[ "foo" ] [ "<!DOCTYPE foo [<!ENTITY bar 'foo'>]><x>&bar;</x>" string>xml children>string ] unit-test
[ T{ xml-chunk f V{ "hello" } } ] [ "hello" string>xml-chunk ] unit-test
[ "1.1" ] [ "<?xml version='1.1'?><x/>" string>xml prolog>> version>> ] unit-test
+[ "ß" ] [ "<x>ß</x>" <string-reader> read-xml children>string ] unit-test
diff --git a/basis/xml/xml.factor b/basis/xml/xml.factor
index 073f46cbae..fba2eafaba 100755
--- a/basis/xml/xml.factor
+++ b/basis/xml/xml.factor
@@ -4,7 +4,8 @@ USING: accessors arrays io io.encodings.binary io.files
io.streams.string kernel namespaces sequences strings io.encodings.utf8
xml.data xml.errors xml.elements ascii xml.entities
xml.writer xml.state xml.autoencoding assocs xml.tokenize
-combinators.short-circuit xml.name splitting io.streams.byte-array ;
+combinators.short-circuit xml.name splitting io.streams.byte-array
+combinators ;
IN: xml
<PRIVATE
@@ -159,6 +160,9 @@ PRIVATE>
xml-stack get first second
] with-state ; inline
+: make-xml ( stream quot -- xml )
+ 0 read-seq make-xml-doc ; inline
+
PRIVATE>
: each-element ( stream quot: ( xml-elem -- ) -- )
@@ -169,14 +173,16 @@ PRIVATE>
] with-state ; inline
: read-xml ( stream -- xml )
- [ start-document [ process ] when* ]
- 0 read-seq make-xml-doc ;
+ dup stream-element-type {
+ { +character+ [ [ check ] make-xml ] }
+ { +byte+ [ [ start-document [ process ] when* ] make-xml ] }
+ } case ;
: read-xml-chunk ( stream -- seq )
[ check ] 1 read-seq <xml-chunk> ;
: string>xml ( string -- xml )
- <string-reader> [ check ] 0 read-seq make-xml-doc ;
+ <string-reader> read-xml ;
: string>xml-chunk ( string -- xml )
<string-reader> read-xml-chunk ;
From fec49cb61683bf2656eecf01fe2a9a3ed2bf83a1 Mon Sep 17 00:00:00 2001
From: Daniel Ehrenberg <littledan@Macintosh-122.local>
Date: Mon, 16 Mar 2009 22:29:38 -0500
Subject: [PATCH 4/8] More expository XML docs
---
basis/xml/traversal/traversal-docs.factor | 18 ++++++++++++++++--
basis/xml/xml-docs.factor | 8 +++++++-
2 files changed, 23 insertions(+), 3 deletions(-)
diff --git a/basis/xml/traversal/traversal-docs.factor b/basis/xml/traversal/traversal-docs.factor
index 1329c4975e..9f26774647 100644
--- a/basis/xml/traversal/traversal-docs.factor
+++ b/basis/xml/traversal/traversal-docs.factor
@@ -1,6 +1,6 @@
! Copyright (C) 2005, 2009 Daniel Ehrenberg
! See http://factorcode.org/license.txt for BSD license.
-USING: help.markup help.syntax xml.data sequences strings ;
+USING: help.markup help.syntax xml.data sequences strings multiline ;
IN: xml.traversal
ABOUT: "xml.traversal"
@@ -8,7 +8,7 @@ ABOUT: "xml.traversal"
ARTICLE: "xml.traversal" "Utilities for traversing XML"
"The " { $vocab-link "xml.traversal" } " vocabulary provides utilities for traversing an XML DOM tree and viewing the contents of a single tag. The following words are defined:"
$nl
- "Note: the difference between deep-tag-named and tag-named is that the former searches recursively among all children and children of children of the tag, while the latter only looks at the direct children, and is therefore more efficient."
+ { $subsection { "xml.traversal" "intro" } }
{ $subsection tag-named }
{ $subsection tags-named }
{ $subsection deep-tag-named }
@@ -20,6 +20,20 @@ ARTICLE: "xml.traversal" "Utilities for traversing XML"
{ $subsection first-child-tag }
{ $subsection assert-tag } ;
+ARTICLE: { "xml.traversal" "intro" } "An example of XML processing"
+"To illustrate how to use the XML library, we develop a simple Atom parser in Factor. Atom is an XML-based syndication format, like RSS. To see the full version of what we develop here, look at " { $snippet "basis/syndication" } " at the " { $snippet "atom1.0" } " word. First, we want to load a file and get a DOM tree for it."
+{ $code <" "file.xml" file>xml "> }
+"No encoding descriptor is needed, because XML files contain sufficient information to auto-detect the encoding. Next, we want to extract information from the tree. To get the title, we can use the following:"
+{ $code <" "title" tag-named children>string "> }
+"The " { $link tag-named } " word finds the first tag named " { $snippet "title" } " in the top level (just under the main tag). Then, with a tag on the stack, its children are asserted to be a string, and the string is returned." $nl
+"For a slightly more complicated example, we can look at how entries are parsed. To get a sequence of tags with the name " { $snippet "entry" } ":"
+{ $code <" "entry" tags-named "> }
+"Imagine that, for each of these, we want to get the URL of the entry. In Atom, the URLs are in a " { $snippet "link" } " tag which is contained in the " { $snippet "entry" } " tag. There are multiple " { $snippet "link" } " tags, but one of them contains the attribute " { $snippet "rel=alternate" } ", and the " { $snippet "href" } " attribute has the URL. So, given an element of the sequence produced in the above quotation, we run the code:"
+{ $code <" "link" tags-named [ "rel" attr "alternate" = ] find nip "> }
+"to get the link tag on the stack, and"
+{ $code <" "href" attr >url "> }
+"to extract the URL from it." ;
+
HELP: deep-tag-named
{ $values { "tag" "an XML tag or document" } { "name/string" "an XML name or string representing a name" } { "matching-tag" tag } }
{ $description "Finds an XML tag with a matching name, recursively searching children and children of children." }
diff --git a/basis/xml/xml-docs.factor b/basis/xml/xml-docs.factor
index 77969c55cd..434209620b 100644
--- a/basis/xml/xml-docs.factor
+++ b/basis/xml/xml-docs.factor
@@ -67,9 +67,9 @@ HELP: string>dtd
ARTICLE: { "xml" "reading" } "Reading XML"
"The following words are used to read something into an XML document"
- { $subsection string>xml }
{ $subsection read-xml }
{ $subsection read-xml-chunk }
+ { $subsection string>xml }
{ $subsection string>xml-chunk }
{ $subsection file>xml }
{ $subsection bytes>xml }
@@ -90,10 +90,16 @@ ARTICLE: { "xml" "events" } "Event-based XML parsing"
{ $subsection pull-event }
{ $subsection pull-elem } ;
+ARTICLE: { "xml" "namespaces" } "Working with XML namespaces"
+"The Factor XML parser implements XML namespaces, and provides convenient utilities for working with them. Anywhere in the public API that a name is accepted as an argument, either a string or an XML name is accepted. If a string is used, it is coerced into a name by giving it a null namespace. Names are stored as " { $link name } " tuples, which have slots for the namespace prefix and namespace URL as well as the main part of the tag name." $nl
+"To make it easier to create XML names, the parsing word " { $snippet "XML-NS:" } " is provided in the " { $vocab-link "xml.syntax" } " vocabulary." $nl
+"When parsing XML, names are automatically augmented with the appropriate namespace URL when the information is available. This does not take into account any XML schema which might allow for such prefixes to be omitted. When generating XML to be written, keep in mind that the XML writer knows only about the literal prefixes and ignores the URLs. It is your job to make sure that they match up correctly, and that there is the appropriate " { $snippet "xmlns" } " declaration." ;
+
ARTICLE: "xml" "XML parser"
"The " { $vocab-link "xml" } " vocabulary implements the XML 1.0 and 1.1 standards, converting strings of text into XML and vice versa. The parser checks for well-formedness but is not validating. There is only partial support for processing DTDs."
{ $subsection { "xml" "reading" } }
{ $subsection { "xml" "events" } }
+ { $subsection { "xml" "namespaces" } }
{ $vocab-subsection "Writing XML" "xml.writer" }
{ $vocab-subsection "XML parsing errors" "xml.errors" }
{ $vocab-subsection "XML entities" "xml.entities" }
From 2f8adf9d4d4941bb51f14a749be635e516af9a39 Mon Sep 17 00:00:00 2001
From: Daniel Ehrenberg <littledan@Macintosh-122.local>
Date: Mon, 16 Mar 2009 22:43:08 -0500
Subject: [PATCH 5/8] Fixing regexp docs
---
basis/regexp/regexp-docs.factor | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/basis/regexp/regexp-docs.factor b/basis/regexp/regexp-docs.factor
index 041c78380b..d62a03ade5 100644
--- a/basis/regexp/regexp-docs.factor
+++ b/basis/regexp/regexp-docs.factor
@@ -22,7 +22,7 @@ ARTICLE: "regexp" "Regular expressions"
{ $subsection { "regexp" "deploy" } } ;
ARTICLE: { "regexp" "intro" } "A quick introduction to regular expressions"
-"Regular expressions are a terse way to do certain simple string processing tasks. For example, to replace all instances of " { $snippet "foo" } " in one string with { $snippet "bar" } ", the following can be used:
+"Regular expressions are a terse way to do certain simple string processing tasks. For example, to replace all instances of " { $snippet "foo" } " in one string with " { $snippet "bar" } ", the following can be used:"
{ $code "R/ foo/ \"bar\" re-replace" }
"That could be done with sequence operations, but consider doing this replacement for an arbitrary number of o's, at least two:"
{ $code "R/ foo+/ \"bar\" re-replace" }
@@ -31,7 +31,7 @@ ARTICLE: { "regexp" "intro" } "A quick introduction to regular expressions"
"To search a file for all lines that match a given regular expression, you could use code like this:"
{ $code <" "file.txt" ascii file-lines [ R/ (f|b)oo+/ re-contains? ] filter "> }
"To test if a string in its entirity matches a regular expression, the following can be used:"
-{ $example <" "fooo" R/ (b|f)oo+/ matches? . "> "t" }
+{ $example <" USING: regexp prettyprint ; "fooo" R/ (b|f)oo+/ matches? . "> "t" }
"Regular expressions can't be used for all parsing tasks. For example, they are not powerful enough to match balancing parentheses." ;
ARTICLE: { "regexp" "construction" } "Constructing regular expressions"
From 9f8ccb67a7955025859f66783968e4edaef555fd Mon Sep 17 00:00:00 2001
From: Daniel Ehrenberg <littledan@Macintosh-122.local>
Date: Mon, 16 Mar 2009 23:49:31 -0500
Subject: [PATCH 6/8] Making disambiguation faster
---
basis/regexp/disambiguate/disambiguate.factor | 45 ++++++++++++++++---
1 file changed, 38 insertions(+), 7 deletions(-)
diff --git a/basis/regexp/disambiguate/disambiguate.factor b/basis/regexp/disambiguate/disambiguate.factor
index 67b1503f9b..876d898cb4 100644
--- a/basis/regexp/disambiguate/disambiguate.factor
+++ b/basis/regexp/disambiguate/disambiguate.factor
@@ -1,7 +1,8 @@
! Copyright (C) 2009 Daniel Ehrenberg.
! See http://factorcode.org/license.txt for BSD license.
USING: kernel accessors regexp.classes math.bits assocs sequences
-arrays sets regexp.dfa math fry regexp.minimize regexp.ast regexp.transition-tables ;
+arrays sets regexp.dfa math fry regexp.minimize regexp.ast
+locals regexp.transition-tables ;
IN: regexp.disambiguate
TUPLE: parts in out ;
@@ -9,7 +10,7 @@ TUPLE: parts in out ;
: make-partition ( choices classes -- partition )
zip [ first ] partition [ values ] bi@ parts boa ;
-: powerset-partition ( classes -- partitions )
+: powerset-partition ( sequence -- partitions )
[ length [ 2^ ] keep ] keep '[
_ <bits> _ make-partition
] map rest ;
@@ -19,19 +20,49 @@ TUPLE: parts in out ;
[ in>> <and-class> ] bi
prefix <and-class> ;
-: get-transitions ( partition state-transitions -- next-states )
- [ in>> ] dip '[ _ at ] gather sift ;
+: singleton-partition ( integer non-integers -- {class,partition} )
+ dupd
+ '[ _ [ class-member? ] with filter ] keep
+ prefix f parts boa
+ 2array ;
+
+: add-out ( seq partition -- partition' )
+ [ out>> append ] [ in>> ] bi swap parts boa ;
+
+: intersection ( seq -- elts )
+ [ f ] [ unclip [ intersect ] reduce ] if-empty ;
+
+: meaningful-integers ( partition table -- integers )
+ [ [ in>> ] [ out>> ] bi ] dip
+ '[ [ _ at ] map intersection ] bi@ diff ;
+
+: class-integers ( classes integers -- table )
+ '[ _ over '[ _ class-member? ] filter ] H{ } map>assoc ;
+
+: add-integers ( partitions classes integers -- partitions )
+ class-integers '[
+ [ _ meaningful-integers ] keep add-out
+ ] map ;
+
+: class-partitions ( classes -- assoc )
+ [ integer? ] partition [
+ dup powerset-partition spin add-integers
+ [ [ partition>class ] keep 2array ] map
+ [ first ] filter
+ ] [ '[ _ singleton-partition ] map ] 2bi append ;
: new-transitions ( transitions -- assoc ) ! assoc is class, partition
values [ keys ] gather
[ tagged-epsilon? not ] filter
- powerset-partition
- [ [ partition>class ] keep ] { } map>assoc
- [ drop ] assoc-filter ;
+ class-partitions ;
+
+: get-transitions ( partition state-transitions -- next-states )
+ [ in>> ] dip '[ _ at ] gather sift ;
: preserving-epsilon ( state-transitions quot -- new-state-transitions )
[ [ drop tagged-epsilon? ] assoc-filter ] bi
assoc-union H{ } assoc-like ; inline
+
: disambiguate ( nfa -- nfa )
expand-ors [
dup new-transitions '[
From 4a79ee9bb92adac7f9d3b118c92dcc97e1e297a1 Mon Sep 17 00:00:00 2001
From: Daniel Ehrenberg <littledan@Macintosh-122.local>
Date: Tue, 17 Mar 2009 00:04:27 -0500
Subject: [PATCH 7/8] Making unicode.data slightly more efficient
---
basis/unicode/data/data.factor | 13 ++++++++-----
1 file changed, 8 insertions(+), 5 deletions(-)
diff --git a/basis/unicode/data/data.factor b/basis/unicode/data/data.factor
index bff4ddeaab..74914e8537 100644
--- a/basis/unicode/data/data.factor
+++ b/basis/unicode/data/data.factor
@@ -5,7 +5,7 @@ io.files hashtables quotations splitting grouping arrays io
math.parser hash2 math.order byte-arrays words namespaces words
compiler.units parser io.encodings.ascii values interval-maps
ascii sets combinators locals math.ranges sorting make
-strings.parser io.encodings.utf8 ;
+strings.parser io.encodings.utf8 memoize ;
IN: unicode.data
VALUE: simple-lower
@@ -108,6 +108,9 @@ CONSTANT: categories
"Zs" "Zl" "Zp"
"Cc" "Cf" "Cs" "Co" }
+MEMO: categories-map ( -- hashtable )
+ categories <enum> [ swap ] H{ } assoc-map-as ;
+
CONSTANT: num-chars HEX: 2FA1E
! the maximum unicode char in the first 3 planes
@@ -124,10 +127,10 @@ CONSTANT: num-chars HEX: 2FA1E
] assoc-each table ;
:: process-category ( data -- category-listing )
- [let | table [ num-chars <byte-array> ] |
- 2 data (process-data) [| char cat |
- cat categories index char table ?set-nth
- ] assoc-each table fill-ranges ] ;
+ num-chars <byte-array> :> table
+ 2 data (process-data) [| char cat |
+ cat categories-map at char table ?set-nth
+ ] assoc-each table fill-ranges ;
: process-names ( data -- names-hash )
1 swap (process-data) [
From a181c220154050ed785fad36a78917f0a21d2dd8 Mon Sep 17 00:00:00 2001
From: Daniel Ehrenberg <littledan@Macintosh-122.local>
Date: Tue, 17 Mar 2009 00:10:55 -0500
Subject: [PATCH 8/8] Regexp docs typo fix
---
basis/regexp/regexp-docs.factor | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/basis/regexp/regexp-docs.factor b/basis/regexp/regexp-docs.factor
index d62a03ade5..6ad340a82d 100644
--- a/basis/regexp/regexp-docs.factor
+++ b/basis/regexp/regexp-docs.factor
@@ -30,7 +30,7 @@ ARTICLE: { "regexp" "intro" } "A quick introduction to regular expressions"
{ $code "R/ (f|b)oo+/ \"bar\" re-replace" }
"To search a file for all lines that match a given regular expression, you could use code like this:"
{ $code <" "file.txt" ascii file-lines [ R/ (f|b)oo+/ re-contains? ] filter "> }
-"To test if a string in its entirity matches a regular expression, the following can be used:"
+"To test if a string in its entirety matches a regular expression, the following can be used:"
{ $example <" USING: regexp prettyprint ; "fooo" R/ (b|f)oo+/ matches? . "> "t" }
"Regular expressions can't be used for all parsing tasks. For example, they are not powerful enough to match balancing parentheses." ;