message: ignored if body empty;

load_thy: removed obsolete dir argument;

abstract type span, tuned interfaces; added report_token, report_span; markup ident tokens;

adapted ThyEdit operations;

added ignored, malformed transitions;

Symbol.source/OuterLex.source: more explicit do_recover argument; scan: pass position; removed obsolete prepare_command (now inlined in isar_syn.ML); renamed prepare_command_failsafe to prepare_command, reports tokens; load_thy: now based on ThyEdit operations, reports tokens and spans; tuned;

Isabelle.command: inline former OuterSyntax.prepare_command; Isar.command: based on fail-safe OuterSyntax.prepare_command;

load thy_edit.ML before outer_syntax.ML;

renamed unknown_span to malformed_span; added ident; tuned;

Symbol.source/OuterLex.source: more explicit do_recover argument;

updated generated file;

rudimentary code setup for set operations

<applet>: more XHTML 1.0 Transitional conformance;

Isar.command: OuterSyntax.prepare_command_failsafe defers syntax errors until execution time;

<pre>: removed xml:space, is already default; result(s): avoid improper nesting of <pre> within <p>;

produce XHTML 1.0 Transitional; tuned charset name; renamed Markup.class to Markup.tclass, to avoid potential conflicts with spacial meaning in markup languages (e.g. HTML);

renamed Markup.class to Markup.tclass, to avoid potential conflicts with spacial meaning in markup languages (e.g. HTML);

Isar.command: do not set position of enclosing transaction, to avoid of clash with the one being prepared here!

changed code setup

re-arranged class dense_linear_order

rudimentary code setup for set operations

moved class wellorder to theory Orderings

added parse_token (from proof_context.ML);

read_tyname/const/const_proper: report position; moved parse_token to syntax.ML; tuned;

pass position to get_fact; tuned;

pass token source to typ/term etc.;

added name property operation; added local_fact; renamed proposition to prop (again);

renamed ML_Lex.val_of to content_of;

unified Args.T with OuterLex.token, renamed some operations; moved thm_sel to attrib.ML;

unified Args.T with OuterLex.token; RESET_VALUE of primitive parsers; export keyword_with; renamed keyword_sid to keyword_ident_or_symbolic; added int (from args.ML); added enum(1)', and_list(1)' (formerly in args.ML); removed obsolete arguments/generic_args1 (cf. parse/parse1 in args.ML);

unified Args.T with OuterLex.token; renamed val_of to content_of; added InternalValue kind; added datatype value/slot with static binding (from args.ML); renamed is_sid to ident_or_symbolic;

unified Args.T with OuterLex.token, renamed some operations; removed obsolete parse_args (cf. parse);

unified Args.T with OuterLex.token, renamed some operations; unified version of thm_sel (formerly in args.ML and spec_parse.ML);

unified Args.T with OuterLex.token; moved datatype value/slot and basic operations to outer_lex.ML; removed unused terminator; removed obsolete !!!, position, nat, int, list(1), enum(1) and_list(1) (cf. outer_parse.ML); removed obsolete thm_sel (cf. attrib.ML); one unified version of parse/parse1 (formerly arguments/generic_args1 in outer_parse.ML);

load args.ML later (after outer_parse.ML);

unified Args.T with OuterLex.token, renamed some operations;

read_asts: report literal tokens;

tuned error message;

pos_of_token: Position.T; removed unused display_token; tuned;

dest: sort strings; report: Output.status;

added literal;

read_token: more robust handling of empty text;

datatype token: maintain range, tuned representation; moved eof, stopper to lexicon.ML;

datatype token: maintain range, tuned representation; tuned messages;

datatype token: maintain range, tuned representation; added eof, stopper (from simple_parse.ML); str_of_token: no special case for EOF; misc tuning;

tuned SymbolPos interface;

YXML.parse: allow text without markup, potentially empty;

added content; simplified implode: interface and permissive padding via Position.distance_of; added range; renamed implode_delim to implode_range;

added distance_of (permissive version); added no_range; tuned;

count offset as well; more uniform treatment of invalid fields; tuned;

added offset/end_offset;

tuned formatting;

clean up dead code

made SML/NJ happy;

FundefLib.try_proof : attempt a proof and see if it works

added lemmas

inner_syntax markup is back;

disabled inner_syntax markup for now;

added read_token -- with optional YXML encoding of position; tuned parse operations: print position instead of echoing input (now encoded!); do not export obsolete read operation;

parse_token: use Syntax.read_token, pass full position information;

tuned;

map_default: more explicit scope;

datatype lexicon: alternative representation using nested Symtab.table;

simplified Antiquote signature;

more precise positions due to SymbolsPos.implode_delim; added source' for SymbolPos.T;

simplified Antiq: regular SymbolPos.text with position; renamed read_arguments to read_antiq; tuned;

renamed SymbolPos.scan_position to SymbolPos.scan_pos; implode/explode: explicit type text = string; added implode_delim; explode: Position.reset_range;

only increment column if valid; more robust handling of invalid entries; clarified line/line_file: no column here; added start = (1, 1); added reset_range; tuned;

install_pp Position.T;

Position.start;

SymbolPos.explode;

improved position handling due to SymbolPos.T;

improved position handling due to SymbolPos.T; SymbolPos.scan_comment; tuned;

Antiquote.read/read_arguments;

updated type of nested sources;

improved position handling due to SymbolPos.T; SymbolPos.scan_comment_body; tuned;

adapted Scan.extend_lexicon/merge_lexicons;

renamed scan_antiquotes to read; renamed scan_arguments to read_arguments; improved position handling due to SymbolPos.T; tuned;

export not_eof;

reorganized lexicon: allow scanning of annotated symbols, tuned representation and interfaces;

advance: single argument (again); added range; tuned;

Symbols with explicit position information.

added General/symbol_pos.ML;

Interpretation command (theory/proof context) no longer simplifies goal.

added lemma

made SML/NJ happy;

Reverted last change, since it caused incompatibilities.

fall back on P.term_group, to avoid problems with inner_syntax markup (due to CodeName.read_const_exprs);

T.end_position_of; tuned;

adapted Antiq;

parse_sort/typ/term/prop: report markup;

sort/typ/term/prop: inner_syntax markup encodes original source position; added typ/term/prop_group (without inner_syntax markup);

removed obsolete range_of (already included in position); added end_position_of; replaced scan_string by scan_quoted (which keeps quotes and includes alt_string as well); misc tuning;

report markup;

Antiq: inner vs. outer position; scan_antiq: use T.scan_quoted;

of_properties: observe Markup.position_properties';

added position_properties'; renamed prop to proposition; added attribute, method;

token: maintain of source, which retains original position information; removed count/counted, advance position via scanned source;

moved OuterLex.count here;

improved recover: also resync on symbol/comment/verbatim delimiters;

advance: operate on symbol list (less overhead);

added token;

fix HOL/ex/LexOrds.thy; add to regression

added report;

removed axiom; added fact, dynamic_fact, prop;

get_fact: report position;

Facts.lookup: return static/dynamic status;

position scanner: encode token range;

added encode_range; tuned;

added end_line, end_column properties;

meta_subst: xsymbols make it work with clean Pure;

abstract type Scan.stopper, position taken from last input token;

abstract type Scan.stopper;

abstract type stopper, may depend on final input;

removed obsolete apply_theorems(_i);

tuned signature; eliminated obsolete pervasives;

removed obsolete note_thms_cmd;

simplified defer_recdef(_i): plain facts via Attrib.eval_thms;

tuned description;

replaced mercurial.cgi by hgwebdir.cgi, resulting in http://isabelle.in.tum.de/repos/isabelle/