langchain_experimental.llms.anthropic_functions.TagParser

class langchain_experimental.llms.anthropic_functions.TagParser[source]

Parser for the tool tags.

A heavy-handed solution, but it’s fast for prototyping.

Might be re-implemented later to restrict scope to the limited grammar, and more efficiency.

Uses an HTML parser to parse a limited grammar that allows for syntax of the form:

INPUT -> JUNK? VALUE* JUNK -> JUNK_CHARACTER+ JUNK_CHARACTER -> whitespace | , VALUE -> <IDENTIFIER>DATA</IDENTIFIER> | OBJECT OBJECT -> <IDENTIFIER>VALUE+</IDENTIFIER> IDENTIFIER -> [a-Z][a-Z0-9_]* DATA -> .*

Interprets the data to allow repetition of tags and recursion to support representation of complex types.

^ Just another approximately wrong grammar specification.

Attributes

CDATA_CONTENT_ELEMENTS

Methods

__init__()

A heavy-handed solution, but it's fast for prototyping.

check_for_whole_start_tag(i)

clear_cdata_mode()

close()

Handle any buffered data.

feed(data)

Feed data to the parser.

get_starttag_text()

Return full source of start tag: '<...>'.

getpos()

Return current line number and offset.

goahead(end)

handle_charref(name)

handle_comment(data)

handle_data(data)

Hook when handling data.

handle_decl(decl)

handle_endtag(tag)

Hook when a tag is closed.

handle_entityref(name)

handle_pi(data)

handle_startendtag(tag, attrs)

handle_starttag(tag, attrs)

Hook when a new tag is encountered.

parse_bogus_comment(i[, report])

parse_comment(i[, report])

parse_declaration(i)

parse_endtag(i)

parse_html_declaration(i)

parse_marked_section(i[, report])

parse_pi(i)

parse_starttag(i)

reset()

Reset this instance.

set_cdata_mode(elem)

unknown_decl(data)

updatepos(i, j)

__init__() None[source]

A heavy-handed solution, but it’s fast for prototyping.

Might be re-implemented later to restrict scope to the limited grammar, and more efficiency.

Uses an HTML parser to parse a limited grammar that allows for syntax of the form:

INPUT -> JUNK? VALUE* JUNK -> JUNK_CHARACTER+ JUNK_CHARACTER -> whitespace | , VALUE -> <IDENTIFIER>DATA</IDENTIFIER> | OBJECT OBJECT -> <IDENTIFIER>VALUE+</IDENTIFIER> IDENTIFIER -> [a-Z][a-Z0-9_]* DATA -> .*

Interprets the data to allow repetition of tags and recursion to support representation of complex types.

^ Just another approximately wrong grammar specification.

Return type

None

check_for_whole_start_tag(i)
clear_cdata_mode()
close()

Handle any buffered data.

feed(data)

Feed data to the parser.

Call this as often as you want, with as little or as much text as you want (may include ‘n’).

get_starttag_text()

Return full source of start tag: ‘<…>’.

getpos()

Return current line number and offset.

goahead(end)
handle_charref(name)
handle_comment(data)
handle_data(data: str) None[source]

Hook when handling data.

Parameters

data (str) –

Return type

None

handle_decl(decl)
handle_endtag(tag: str) None[source]

Hook when a tag is closed.

Parameters

tag (str) –

Return type

None

handle_entityref(name)
handle_pi(data)
handle_startendtag(tag, attrs)
handle_starttag(tag: str, attrs: Any) None[source]

Hook when a new tag is encountered.

Parameters
  • tag (str) –

  • attrs (Any) –

Return type

None

parse_bogus_comment(i, report=1)
parse_comment(i, report=1)
parse_declaration(i)
parse_endtag(i)
parse_html_declaration(i)
parse_marked_section(i, report=1)
parse_pi(i)
parse_starttag(i)
reset()

Reset this instance. Loses all unprocessed data.

set_cdata_mode(elem)
unknown_decl(data)
updatepos(i, j)