#!/usr/bin/perl # # Revision History: # ---------------- # Revision 0: Implemented gremlins, hyphenation, sectioning, # initial stab at underscore-handling. # Revision 1: Swapped \w for \S in underscore regexps and added # checks to for punctuation and EOL-terminated underscores # Revision 2: Initial attempt at underscores. # Revision 3: Finished work on underscores. # Added URLs. Defined internal latex commands # Revision 4: More work on URLs (brackets now added by latex commands) # Sketched out skeleton of remaining implementation # Adds latex header/footer and sets title # Revision 5: Added itemization # Revision 6: Smart quotes, hrules # Fixed (somewhat) the URL block situation (URL vs. inline URL # context now noted). # Revision 7: URLs now separated from paragraphs and encapsulated in tabulars # Revision 8: Cleaned up argument processing # Added -mdj option to grab `author' # Moved URL blocks back out of tables (now using \\ + \noindent # to do grouping) # Revision 9: Fixed (broken) argument processing from r8 # Added -c flag for writing to stdout # Will also guess unspecified outfile name for .etx files # Revision 10: Misc. cleanup, will do more guessing w/r/t filenames, # New symbol substitution for abbrevs, exponents # Revision 11: Fixed end-of-line hyphens when word on next line # terminates with punctuation # MDJ-mode tweaking for boilerplate and auto-activation # based on document's title # Now handles backslashes in source text # Revision 12: Added two special cases for (text_) and [text_] conditions # Further messes with URLs in boilerplate section # # Revision 13: Complete Reorganization. Now each pass dealt with within # its own subroutine # Revision 14: Moved url display over to using url.sty, still not working... # 14b: Moved back from url.sty to hyperref # # Revision 15: Completed switch to hyperref, modified url functions to allow # for easier line-wrapping # # Revision 16: Added switch to disable fancy footers, cleaned up defs for public use # Revision 17: Fixed end-of-line hyphenation to look for \s(\w+-)+$ which # now catches e.g., "multi-word- # hyphenation-chains" # # Revision 18: Now uses the relsize package for tag-related size changes. # Converts pgp signature to monospaced font # Revision 19: Another hyphenation fix for words in the chain containing punctuation # # Revision 20: Added `next' calls to skip special formatting within \etxURL{} and \etxLink{} lines # Revision 21: The manual-hyphenation code was grabbing ' -$' runs missing that they're supposed to be dashes # Revision 22: Properly space "` and '" within-quotation quotes # Revision 23: Redo manual hyphenation fix. Now requires that the first character in # a manually-hyphenated chain be a word char (so "--$" isn't considered # to be a hyphenated "word") # Revision 24: `final' (?) change to hyphenation. Fixes spacing to turn "foo -$" into # "foo---bar$" # Revision 25: Smart-quotes: now counts: {" as an opening quote # Revision 26: Use ~ for spacing after titles like: Mr. Dr. Ms. Mrs. # Use smallcaps in mdj for `US$' # Convert ... to ldots # Revision 27: Undid first-char requirement from rev 23 b/c it was missing things like # `` (pre-$ '' and `` "under-$ ''. Besides, we have a special case now for # lines ending ``--$'' # Revision 28: Realized the much-tweaked hyphen rule wasn't correctly grabbing the # whitespace on the second line. Fixed now. # Revision 29: Added Acronym-to-SmallCaps conversion # Changed \\[-1ex] to \\*[-1ex] for spacing after url lines # Revision 30: SmallCaps bug fixes: avoids verbatim environments # catches things that start with only 2 caps (e.g., CD-ROM) # Substitutes textasciitilde for ~ in visible text (i.e., not hrefs) # Revision 31: Changes to \parskip in itemize environments and linebreak size # after non-inline urls # Revision 32: Smarter tex escape filtering when counting characters in potential # section/subsection lines # Revision 33: Cleaned up TLA regex code so we don't need duplicate lines anymore # Added switch to suppress TLA conversion and updated help message # Revision 34: Since \emph{\textsc{ * }} doesn't exist, change to just \emph. Updated # .* in regexp to be non-greedy, so it catches the `}}' at the end. # Revision 35: Add digits to TLA regex so things like MP3 and 32X get caught # # Revision 36: Major change to url-handling. Groups of urls now use the urlblock and linkblock # environments for formatting. This means better line-wrapping of long urls, more # reliable spacing, and fewer underfull-hbox warnings. Also, the itemize environment # has been changed to the briefs environment (to give us a bit more control over # line spacing). Relocated \parskip setting from doc body to brief environment. # Revision 37: Fix to TLA+# regex behavior. Since it considers 3+digit numbers to be caps runs, # add check at end of conversions to convert-back numbers (e.g., \textsc{2002} -> 2002) # Revision 38: Added multiples-handling to TLAs (e.g, `32X') # Revision 39: Made smallcaps `1' fix look for multiple `1' chars within the \textsc{} # Improved(?) fix for smallcaps/italic conversion # Revision 40: Undid smallcaps `1' fix and changed the bembo interpretation of SC 1 to suit my needs. # Had another go at fixing smallcaps/italic problems. # Revision 41: Fixed custom acronyms conversion to deal with escapes proparly (e.g. `AT\&T') # Added warning when more than one hrule is found in an MDJ document # Revision 42: Underscores now catch compound words where only the first portion is to be # emphasized (e.g. `_non-_PostScript') # Emboldens Q & A labels in ask-the-staff sections # Revision 43: Fixes carets that aren't being used to denote exponents # Revision 44: Added handlers for Vs. v. and Q&A # Revision 45: Added quick & dirty line wrapping for long section titles (char counting) # Revision 46: Cleaned up public frontmatter (so it actually compiles now) # Revision 47: Removed requirement that vs. be followed by a cap for spacing. # Revision 48: Added TidBITS mode for cleaning up ToC and erasing ads # Revision 49: Now distinguishes between 1/2 columns and landscape/portrait orientation # via separate switches. Adjusts column spacing for portrait 2-col mode. # Revision 50: In non-mxj docs, looks for double-dollar tag (meaning end-of-doc) and sets all # following lines as boilerplate text # Revision 51: Tweaks to the margin sizes in the two-column portrait layout # Revision 52: Fixed bug in backslash handling that could introduce spaces and newlines(!) # Revision 53: Fixed quoted-backslash bug in preprogrammed acronyms # Revision 54: Fixed a typo in the gremlin zapping that missed '\{' # Revision 55: Migrated doc output into its own subroutine (for neatness's sake) # Made footer start on first page rather than second # Revision 56: Updated the hyphenation behavior in URLs, now tries to break at # slashes or various other punctuation chars rather than at every char # Revision 57: Disregard the [9] in [9][X] when checking for URL refs # Revision 58: Misc. tweaks to the hyphenation changes # Revision 59: Now correctly calls \textdollaroldstyle for dollar signs in OsF # Revision 60: Changed the layout to make margins smaller # Revision 61: Fixed exponentiation (added braces and switched to superior numerals) # Revision 62: Added custom hyphenation for words with InterCapping # Revision 63: Added margin-kerning for non-public mode # Revision 64: Added boxing of [9] and [X] in product showcase # Revision 65: Added switch to select placing the version boxes inline w/i title # Revision 66: Fixed bug where the first line of the input file was being discarded # Switched textoldstyledollar back to a smallcap dollar (since nobody includes it) # Revision 67: Broke up latex frontmatter-creation into separate subroutines # Smart-rewrapping of \section titles now actually compiles the latex # code and measures the resulting text rather than just counting chars. # Fixed handling of nested quote: "' (was previously looking for "`) # Fixed one of the isPublic latex headers which was missing # Revision 68: Updated the \section breaking to split lines on ; and - as well as : # Revision 69: Added + to legal dividers in small cap blocks (in addition to - and / ) # Revision 70: Changed -scan option to just print the table of contents, not produce output # Revision 71: Fixed a condition where a line ending with `-' would have the last word # removed, but not re-added if the next line was totally empty. # Updated the non-public headers to reflect texmf changes # Revision 72: Added a (hopefully) temporary hack to smallcap (not italicize) MDJ_ # Also fixed quote-education for "s at the beginnings of lines # Revision 73: Now filters out blanks at the end of section- or article-title lines # so it doesn't misidentify --- as an hrule # Revision 74: The hack from rev72 was modifying a temporary string instead of the real one... # Revision 75: Fixed side-effect of using \, as space around em-dashes. Now consider \, as # white-space in quote-education phase # Revision 76: Massively sped-up the small-capification of custom acronyms by breaking it out # of its loop and pattern-matching using s///gm. Turned off smart-testing of line- # breaking of title text if this is a public version (since other users may not # have TeX in a working state when they run the script). # Revision 77: Streamlined the rest of the small-capping code. # Revision 78: Fix to undo small-capping within URLs # Revision 79: Some text-size tweaks. # Revision 80: Did some reformatting of the MDJ boilerplate section and added [D] symbol # Now reads .txt files as well as .etx # Revision 81: Added table of contents and fixed urlblock problem choking on urls on the # last line of text # Revision 82: Changed the period spacing after a `vs.' ToDo: update the others too # Revision 83: Added period spacing for `Inc.' # Changed Paragraphs starting with **subhead** -- to use \topicpar{} # Fixed the [9] and [X] boxes for product showcase entries # Tightened spacing in link/urlblocks (for non-public version) # Revision 84: Made the [9] and [X] boxes not appear in TOC lines (latex didn't like that) # $$ $SetexRevisionNumber=84; # # Tag Equivalence: # bold-tt *text* \etxBold # underline-tt _text_ \etxUline # hot-tt text_ \etxEmph # title-tt ===== \etxTitle # ?? ----- \section # **\w** \subsection # [\d] \etxRef # <*> \etx(Inline)URL # [\d] <*> \etxLink # ########################################################################## $isPublic=1; # If 0, output is dependent on my local setup my(@data, # The contents of the input file $filename, # The file we'll write our output to $opt_twoup, # Flag for 2up layout $opt_numCols, # Number of columns (by default: 2 for landscape, 1 for portrait) $opt_nofooter, # Flag to suppress printing fancy footer $opt_nosmallcaps, # Flag to suppress converting caps words to smallcaps $opt_mdj, # Flag for special mdj-handling $opt_tidbits, # Flag for special tidbits-handling $opt_printToC, # Flag to print a table of contents as file is processed $opt_verbose); # Flag to be chatty my($title_linenum, $title_text, @customHyphenations); &readInputFile; ## Parse Command Line & read in file &zapGremlins; ## Escape all problematic special-characters # and set up custom \hyphenation patterns &doTidbitsPreprocessing; &doSectioning; ## Find document title (set $title_linenum/text) # and divide into sections based on --- and === # tags &doHyphenation; ## Extend `-' chars as needed and do line-wrapping &doMdjConversions; &doSmallCapping;# Convert TLAs to small caps &processStars; # Convert ** tag to embolden &processUnderscores; # Try to infer _italicization_by_underscore_ &processLinks; # Convert to use \etxUrl{} or \etxLink{} &doItemization; # Convert lines with leading `*'s to \item bullets &doCleanup; &packageDocument; # Piece together the latex header, footer, body, etc. &outputDocument; # Write the text to file exit(0); ## Done ######################################################### # # readInputFile() # # Parses command line options and reads in the input file # # Reads Variables: @ARGV # Sets Variables: $opt_mdj Boolean: Mdj-optimization # $opt_twoup Boolean: Twoup layout # $filename String: Name of file to write output to # @data Array: The lines of the input file sub readInputFile{ my($in_filename,$out_filename); # Process args (setting flags, etc.) for(@ARGV){ if (m/^-/){ if (m/^-1/){ $opt_numCols=1; } elsif (m/^-2/){ $opt_numCols=2; } elsif (m/^-l/){ $opt_twoup=1; if($numCols==0){ $numCols=2; }} elsif (m/^-m/){ $opt_mdj=1; } elsif (m/^-t/){ $opt_tidbits=1; } #elsif (m/^-t/){ $opt_printToC=1; } elsif (m/^-c/){ $out_filename="__StdOut__"; } elsif (m/^-nof/) { $opt_nofooter=1; } elsif (m/^-nos/) { $opt_nosmallcaps=1; } elsif (m/^-v/){ $opt_verbose=1; } elsif (m/^-s/){ $opt_printToC=1; } elsif (m/^--v/){ die "seTeX version $SetexRevisionNumber\n"; } elsif (m/^--h/){ print "Convert setext-formatted documents to latex markup.\n\n"; print "Usage: setex [--help] [-c] [-2] [-landscape] [-mxj | -tidbits] [-nofoot] [-nosc] [-v] [-scan] [output file]\n\n"; print "Options:\n"; print " -c Write to stdout\n"; print " -2 Print in two-column layout\n"; print " -landscape Print in landscape orientation\n"; print " -mxj Use special MDJ/MWJ-handling\n"; print " -tidbits Use special TidBITS-handling\n"; print " -nofoot Don't print a footer\n"; print " -nosc Don't convert CAPS to smallcaps\n"; print " -scan Print out a table of contents listing\n"; exit; } } elsif(! $in_filename){ $in_filename=$_; } elsif (! $out_filename){ $out_filename=$_; } else{ print "Mystery argument: $_\n"; } } ## Verify existence of input file if ($in_filename eq ""){ die("Usage: setex [--help] [-c] [-2] [-landscape] [-mxj | -tidbits] [-nofoot] [-nosc] [-v] [-scan] [output file]\n"); }elsif(! -f $in_filename){ my $alt_filename = join('',$in_filename,".etx"); if (! -f $alt_filename){ die("File: $in_filename doesn't exist\n"); }else{ $in_filename = $alt_filename; } } ## Deal with undefined $out_filename if(! $out_filename && $in_filename =~ m/[[:print:]]*.etx/){ $out_filename = `basename $in_filename .etx`; chomp($out_filename); $out_filename = "$out_filename.tex"; }elsif(! $out_filename && $in_filename =~ m/[[:print:]]*.txt/){ $out_filename = `basename $in_filename .txt`; chomp($out_filename); $out_filename = "$out_filename.tex"; }elsif(! $out_filename){ die("No output file specified.\n"); } ## Deal with potentially clobbering output file if ($out_filename ne "__StdOut__" && -f $out_filename){ print STDERR "File \"$out_filename\" exists. Overwrite? (Y/n) "; $_=; if ( !(m/[Yy]/ || length($_)==1) ){ die("Okay, quitting.\n"); } } ## Read in the input file open(INFILE, "<$in_filename") || die("Can't open input file: $!"); while (){ push @data, $_; } close(INFILE); chomp(@data); ## Return our values $filename = $out_filename; push @retlist, $opt_twoup; push @retlist, $opt_mdj; push @retlist, $out_filename; return @retlist } ################################################################### # # zapGremlins() # # Escapes all the special laTeX characters # # Modifies: @data # @customHyphenations # sub zapGremlins{ $opt_verbose && print STDERR "zapGremlins"; my(%hyphHash); # Assumes @data is the text we're working on for ($i=0; $i<= $#data; $i++){ $_ = $data[$i]; # $_ is default for m// # zap $ s/\$/\\\$/g; s/\{/\\\{/g; $data[$i] = $_; # zap { s/\}/\\\}/g; $data[$i] = $_; # zap } # zap any \ we didn't create in the `$', `{', or `}' steps s/\\(?![\$\}\{])/\\textbackslash{}/g; $data[$i] = $_; # Zap remaining gremlins s/\%/\\\%/g; $data[$i] = $_; # zap % s/\#/\\\#/g; $data[$i] = $_; # zap # s/\&/\\\&/g; $data[$i] = $_; # zap & # zap ~ if(m/~\s/ && ($` !~ //) ){ s/~\s/\\textasciitilde\\ /g; $data[$i] = $_; }elsif (m/~.*(?!\>)/ && ($` !~ //) ){ s/~/\\textasciitilde /g; $data[$i] = $_; } ## Find words with InterCapping for custom hyphenizations while ($data[$i] =~ m/([A-Za-z]+[a-z][A-Z][A-Za-z]+)/g){ $theWord = $1; $hyphenatedWord = $1; $hyphenatedWord =~ s/([a-z])([A-Z])/$1-$2/g; # Store it in a hash so we can weed out duplicates $hyphHash{"$theWord"} = $hyphenatedWord; }; } # Save all the unique interCapped words we found for later... @customHyphenations = values %hyphHash; $opt_verbose && print STDERR ".\n"; } ################################################################### # # doHyphenation() # # Translates ascii `-' characters to proper dash lengths based on # line context # # Modifies: @data # sub doHyphenation{ $opt_verbose && print STDERR "doHyphenation"; my ($carried_word); for ($i = 0; $i <= $#data; $i++) { $_ = $data[$i]; # $_ is default for m// # There are 4 different conditions under which we # need to deal with dashes: # a) long dividers: " -- " -> " ---" # b) short dividers: " - " -> " -- " # c) end-of-line: " -$" -> " --$" # d) manually hyphenated words: "\w-$" ## Case a: s/ -- / --- /g; # Usually a spacer, so leave the whitespace ## Case b: s/ - /---/g; # Typically used midsentence as a dash ## Case c: s/ -$/---/g; # then use rule D to reflow the lines $data[$i] = $_; # Make sure we ignore things that aren't hyphens/dashes # (i.e., bail out if we shouldn't handle case D and rewrap) next if ( m/----+$/ || m/ -+$/); ## Case d: ###s/\s((\w[\w[:punct:]]+-)+)$// && do{ ## clip out hyphenated baseword ### ^__ At lease first char must be alphanum if($data[$i+1] =~ m/^(\s*)(\w*)(\W)/){ # ^-- make sure there's text on the next line, since otherwise this is # probably not a hypenation and we should probably leave it as is s/\s(([\w[:punct:]]+-)+)$// && do{ ## clip out hyphenated baseword #print "$filename $i:\t\"$data[$i]\"\n"; $data[$i] = $_; ## save clipped line $carried_word=$1; #print "$filename $i:\t\"$_\": \"$carried_word\"\n"; #print "\t\t\"$_\"\n"; $_ = $data[$i+1]; # insert baseword at front of next line: s/^(\s*)(\w*)(\W)/$1$carried_word$2$3/; #print "\t\t\"$_\"\n"; $data[$i+1] = $_; ## save next line }; } } $opt_verbose && print STDERR ".\n"; } ################################################################### # # doSectioning() # # Looks for lines of --- or === characters marking section or # document titles respectively. Also, catch text flanked by # **double stars** as being a subsection. # # Modifies: @data # Sets Variables: $title_linenum # The line with the document title # $title_text # The Text of the document title # # Note: may set $opt_mdj to 1 if the document title looks like an MDJ issue sub doSectioning{ $opt_verbose && print STDERR "doSectioning"; my ($title_len, $sect_len, $prev_len, $prev_line, $new_line,@table_of_contents); $title_linenum = -1; for ($i = 0; $i <= $#data; $i++) { $_ = $data[$i]; # $_ is default for m// ## Find Document Title (single line with `====' on following line) /(^==+)[\s]*$/ && do{ ## If the number of `='s is same as num chars on prev line, # this is the title. But first screen out any `\' escapes # before calculating the previous line's length $title_len=length($1); # filter tex commands/escapes $prev_line = $data[$i-1]; $prev_line =~ s/(?:\\[a-z]+\{)+(.*)/$1/g; $prev_line =~ s/(?= 22 && &checkLineBreak('section',"$data[$i-1]")){ $opt_verbose && print STDERR "o"; $new_line =~ s/(:|;| - )/$1\\\\/; # If title too long, wrap at first colon $new_line =~ s/section{/section[$data[$i-1]]{/; }else{ $opt_verbose && print STDERR "."; } if ($opt_printToC && $data[$i-1] !~ m/Top of the Day/i){ #print STDERR "$data[$i-1]\n"; push @table_of_contents,"$data[$i-1]\n"; } @data[$i-1,$i] = $new_line; $i--; } }; ## Find Subections (lines starting & ending with `**') /^\*\*(.+)\*\*\s*$/ && do{ $new_line="\\subsection\{$1}"; ## Special case, check for ** tags within subsection title if ($new_line =~ m/^(.*)\*\*(.+)\*\*(.*)$/){ $new_line = "$1\\etxBold\{$2\}$3\n"; } if ($opt_printToC){ #print STDERR " - $1\n"; push @table_of_contents," - $1\n"; } $data[$i]=$new_line; # Update line in doc }; ## Find SubSubsections (lines starting with a '** **' complex, # but continuing on into a paragraph) if (! $isPublic){ $data[$i] =~ s/^\*\*(.+)\*\*\s*?--(.*)$/\\topicpar\{$1\} $2/; } $data[$i] =~ s/^\*\*(.+)\*\*(\s.*)$/\\topicpar\{$1\}$2/; } $opt_verbose && print STDERR ".\n"; # Dump the listing to stderr if (@table_of_contents){ unshift @table_of_contents, "$title_text\n"; print STDERR "\n@table_of_contents\n"; exit(0); } } ################################################################### # # doTidbitsPreprocessing() # # Reads: $opt_tidbits # Modifies: @data sub doTidbitsPreprocessing{ if ($opt_tidbits){ $opt_verbose && print STDERR "doTidbitsConversions"; # Strip out the ads my($begin_ads,$end_ads,$begin_clip); $begin_ads=-1,$end_ads=-1,$begin_clip=-1; for ($i = 0; $i <= $#data; $i++) { if ($begin_clip==-1 && $data[$i] =~ m|^0 && $blockstart<$blockend && ($blockend-$blockstart)<16){ $data[$blockend] = join('',"$data[$blockend]","\\end{verbatim}"); $data[$blockstart] = join('',"\\begin{verbatim}","$data[$blockstart]"); $blockstart=-1, $blockend=-1; } } ## Find Q & A blocks for ($i = 0; $i <= $#data; $i++) { $data[$i] =~ s/^Q/\\noindent\\textsb{Q}/; $data[$i] =~ s/^A/\\vspace{2ex plus0ex minus2ex}\\noindent\\textsb{A}/; } ## Do misc. MDJ tweaks for ($i = 0; $i <= $#data; $i++) { $_ = $data[$i]; # $_ is default for m// # Make copyright line the `author' for beauty's sake if (!$have_set_author && m/^\s*(Copyright.*)$/){ $data[$i]="\\etxAuthor{$1}"; $have_set_author=1; } # Clip out their weird pasting artifacts if (m/ v$/ && $data[$i+1] =~ m/^(\W)*\\s*$/ || # skip links $data[$i] =~ m/^\s*\<([[:print:]]+?)\>\s*/ || # skip urls $data[$i] =~ m/^\\etxTitle/ || # skip title line (in sans) $data[$i] =~ m/-{4,}/ || $isVerb != 0); # If we didn't skip, then it's okay to apply smallcaps operation on this line $data[$i] =~ s/\b([A-Z0-9\-\+\/\\&]{3,})s\b/\\textsc{\L$1}s/g; # Catch caps words ending with `s' $data[$i] =~ s/\b([A-Z0-9\-\+\/\\&]{3,})\b/\\textsc{\L$1}/g; # Catch caps words $data[$i] =~ s/\b([A-Z][0-9])\b/\\textsc{\L$1$2}/g; # Catch 2-char strings like 3D $data[$i] =~ s/\b([0-9][A-Z])\b/\\textsc{\L$1$2}/g; # Catch 2-char strings like G4 # Handle underscored caps too $data[$i] =~ s/(?<=[_ [:punct:]])([A-Z0-9\-\+\/&\\]{3,})(?=[_ [:punct:]])/\\textsc{\L$1}/g; ### Undo side-effect of... { #...small-capping 3+digit numbers while ($data[$i] =~ m/\\textsc{([0-9\-\\\/]+)}/){ $data[$i] =~ s/\\textsc{([0-9\-\\\/]+)}/$1/g; } #...single caps near dashes $data[$i] =~ s/\\textsc{(-+[a-z])}/\U$1/g; $data[$i] =~ s/\\textsc{([a-z]-+)}/\U$1/g; #...smallcapping 68K (which looks bad) $data[$i] =~ s/\\textsc{68k}/68K/; } } # Catch any 2-letter abbrevs specified in the @acronyms array my @acronyms = qw(CD XP IP TV ID PC US AM PM MX GX VP HP AG); my $wholeDoc = join("\n",@data); my $bspace='(?<=[ [:punct:]])'; # matches a divider char Before pattern my $fspace='(?=[ [:punct:]])'; # matches a divider char Following pattern foreach $ac (@acronyms){ my $lc_ac = lc $ac; # Downcase the acronym $lc_ac =~ s/\\\\/\\/g; # Deal with quoting of backslash chars $wholeDoc =~ s/$bspace${ac}$fspace/\\textsc{$lc_ac}/mg; $wholeDoc =~ s/$bspace${ac}s$fspace/\\textsc{$lc_ac}s/mg; $wholeDoc =~ s/$bspace${ac}$/\\textsc{$lc_ac}/mg; $wholeDoc =~ s/$bspace${ac}s$/\\textsc{$lc_ac}s/mg; } ### Undo side-effect of small-capping within a URL $wholeDoc =~ s/(<.*?)\\textsc\{(.*?)\}(.*>)/$1\U$2$3/mg; # Commit changes @data = split(/\n/,$wholeDoc); $opt_verbose && print STDERR ".\n"; } ################################################################### # # processStars() # # Modifies: @data # # Emboldens words with bold-tt tags (e.g., *bold words*) sub processStars{ $opt_verbose && print STDERR "processStars"; my ($new_line); for ($i = 0; $i <= $#data; $i++) { $_ = $data[$i]; # $_ is default for m// # It's unlikely we'll see an asterisk in a url, but just # to be on the safe side we'll skip the line... next if ( m/^\s*\[(\d+)\]\s+\<([[:print:]]+?)\>\s*$/ || m/^\s*\<([[:print:]]+?)\>\s*/ ); ## Process Double-Stars while ( m/(^.*\s)(\*\*)([\w\s[:punct:]]*)(\*\*)(.*)$/g ){ # $2 and $4 are the junk *'s $new_line = join('',$1,"\\etxBold{",$3,"}",$5); $_ = $new_line; $data[$i] = $_; } ## Process Stars while ( m/(^.*\s)(\*)([\w\s[:punct:]]*)(\*)(.*)$/g ){ # $2 and $4 are the junk *'s $new_line = join('',$1,"\\etxBold{",$3,"}",$5); $_ = $new_line; $data[$i] = $_; } } $opt_verbose && print STDERR ".\n"; } ################################################################### # # processUnderscores() # # Modifies: @data # # Converts underscores representing hot-tt or underline-tt, or # escapes them if they're embedded in the middle of a_word. # sub processUnderscores{ $opt_verbose && print STDERR "processUnderscores"; my($pre, $post, $clipping, $new_line); for ($i = 0; $i <= $#data; $i++) { $_ = $data[$i]; # $_ is default for m// # We want to leave underscores untouched in urls, so bail out # if this line is either an \extLink{} or \etxURL{} respectively. next if ( m/^\s*\[(\d+)\]\s+\<([[:print:]]+?)\>\s*$/ || m/^\s*\<([[:print:]]+?)\>\s*/ ); #!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! # This is a nasty hack that should probably be moved somewhere # more appropriate (that or hot_ tags should be set properly # to begin with). In any case, we want to prevent MDJ_ from # being italicized instead of smallcapped. s/\\textsc\{mdj\}_/\\textsc\{mdj\}/g; s/\\textsc\{mwj\}_/\\textsc\{mwj\}/g; #!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ## Process underscores # # There are two varieties of underscore tag: # a) Underlining: _the_text_ # b) Emphasizing: the_text_ # # c) Degenerate underlining(?): _the_text # #################### ## Case a: while ( m/(^.*\s)(_(?:\S+_)+)([\s[:punct:]].*)$/g ){ # look for whitespace_text_whitespace $pre=$1; $post=$3; $clipping = substr($2,1,length($2)-2); $clipping =~ s/_/ /g; $clipping = join('',"\\etxUline{",$clipping,"}"); $new_line = join('',$pre,$clipping,$post); $_=$new_line; $data[$i] = $_; } while ( m/(^.*[\s[:punct:]])(_(?:\S+_)+)([\s[:punct:]].*)$/g ){ # As above but consider punctuation to be whitespace too $pre=$1; $post=$3; $clipping = substr($2,1,length($2)-2); $clipping =~ s/_/ /g; $clipping = join('',"\\etxUline{",$clipping,"}"); $new_line = join('',$pre,$clipping,$post); $_=$new_line; $data[$i] = $_; } while ( m/(^.*\s)(_(?:\S+_)+)$/g ){ ## look for _text[eol] $pre=$1; $post=$3; $clipping = substr($2,1,length($2)-2); $clipping =~ s/_/ /g; $clipping = join('',"\\etxUline{",$clipping,"}"); $new_line = join('',$pre,$clipping,$post); $_=$new_line; $data[$i] = $_; } #################### ## Case b: while (m/(^.*\\\w+\{)((?:\S+_)+)(\s.*)$/g){ # look for \word{text_WHITESPACE (b/c we need to make sure we're not stepping over previously applied latex commands $pre=$1; $post=$3; $clipping = substr($2,0,length($2)-1); $clipping =~ s/_/ /g; $clipping = join('',"\\etxEmph{",$clipping,"}"); $new_line = join('',$pre,$clipping,$post); $_=$new_line; $data[$i] = $_; } while (m/(^.*[\s\(\'\"])((?:\S+_)+)(\s.*)$/g){ # look for whitespace|text_whitespace $pre=$1; $post=$3; $clipping = substr($2,0,length($2)-1); $clipping =~ s/_/ /g; $clipping = join('',"\\etxEmph{",$clipping,"}"); $new_line = join('',$pre,$clipping,$post); $_=$new_line; $data[$i] = $_; } while (m/(^.*\()((?:\S+_)+)(\).*)$/g){ ## Catch punctuation terminator (but look for pairs of parens) ## i.e., look for (text_) $pre=$1; $post=$3; $clipping = substr($2,0,length($2)-1); $clipping =~ s/_/ /g; $clipping = join('',"\\etxEmph{",$clipping,"}"); $new_line = join('',$pre,$clipping,$post); $_=$new_line; $data[$i] = $_; } while (m/(^.*\[)((?:\S+_)+)(\].*)$/g){ ## Catch punctuation terminator (but look for pairs of brackets) ## i.e., look for [text_] $pre=$1; $post=$3; $clipping = substr($2,0,length($2)-1); $clipping =~ s/_/ /g; $clipping = join('',"\\etxEmph{",$clipping,"}"); $new_line = join('',$pre,$clipping,$post); $_=$new_line; $data[$i] = $_; } while (m/(^.*[\s\(\'\"])((?:\S+_)+)([\s[:punct:]].*)$/g){ ## Catch punctuation terminator # As above but consider general punctuation to be whitespace too $pre=$1; $post=$3; $clipping = substr($2,0,length($2)-1); $clipping =~ s/_/ /g; $clipping = join('',"\\etxEmph{",$clipping,"}"); $new_line = join('',$pre,$clipping,$post); $_=$new_line; $data[$i] = $_; } while (m/(^.*[\s\(\'\"])((?:\S+_)+)$/g){ ## look for whitespace|text_[eol] $pre=$1; $clipping = substr($2,0,length($2)-1); $clipping =~ s/_/ /g; $clipping = join('',"\\etxEmph{",$clipping,"}"); $new_line = join('',$pre,$clipping); $_=$new_line; $data[$i] = $_; } while (m/(?<=\s)(_(?:\S+_)+)/g){ $pre="$`"; $post="$'"; $clipping = substr($1,1,length($2)-1); $clipping =~ s/_/ /g; $clipping = join('',"\\etxEmph{",$clipping,"}"); $new_line = join('',$pre,$clipping,$post); $_=$new_line; $data[$i] = $_; } ## Zap remaining (gremlin) underscores s/_/\\_/g; # Escape each _ $data[$i] = $_; # (But then unzap anything within a URL) #while (m/^(.*\<.*)\\(.*\>.*)$/){ # $_ = "$1$2"; #} #$data[$i] = $_; } $opt_verbose && print STDERR ".\n"; } ################################################################### # # processLinks() # # Modifies: @data # # Deals with hyperlinks and footnotes within the document. URLs # are defined by being between and footnote # references are digits within square brackets (e.g., [2]) # sub processLinks{ $opt_verbose && print STDERR "processLinks"; for ($i = 0; $i <= $#data; $i++) { $_ = $data[$i]; # $_ is default for m// ## There are four types of URL/Reference linking # # Alone: The only non-whitespace characters on the \etxUrl{} # line are between the angle brackets # Referenced: Angle bracket-enclosed text is preceded by \etxLink{} # a hard-bracketed reference number # # Inline Ref: Hard-bracketed number with non-whitespace \etxRef{} # characters on one or more sides # Inline URL: Angle-bracketed text with non-whitespace \etxInlineUrl{} # characters on one or more sides # Check for paired URL if ( m/^\s*\[(\d+)\]\s+\<([[:print:]]+?)\>\s*$/ ){ ## Referenced URL # $1: the ref # $2: the URL my ($ref_number, $plain_url, $padded_url); $ref_number=$1; $plain_url=$padded_url=$2; #$padded_url =~ s/(?\s*$/){ my ($plain_url, $padded_url); # Convert to \etxUrl{} $plain_url=$padded_url=$1; #$padded_url =~ s/(?\s*/\\etxUrl{$plain_url}{$padded_url}/g; # Add trailing linebreaks as needed # if (! ($data[$i+1] =~ m/^(\W)*$/ || $data[$i+1] =~ m/^$/) ){ # #if (!($data[$i+1] =~ m/\^(\W)*$/)){ # $_ = "$_ \\\\* "; # }else{$_ = "$_ \\\\*[-1.5ex] ";} } $data[$i] = $_; ## Replace inline URLs if (m/\<([[:print:]]+?)\>/){ my ($plain_url, $padded_url); $plain_url=$padded_url=$1; #$padded_url =~ s/(?/\\etxInlineUrl{$plain_url}{$padded_url}/g; } $data[$i] = $_; } } $opt_verbose && print STDERR "."; ## Walk through doc finding blocks of URLs my(@url_begin, @url_end, $in_url_region); $in_url_region = 0; # (i.e., not in an itemize region) for ($i = 0; $i <= $#data; $i++) { $_ = $data[$i]; # $_ is default for m// if (m/\\etxUrl/ || m/\\etxLink/){ if (! $in_url_region){ ## This is the beginning of a region push @url_begin, $i; $in_url_region=1; # If this is the last line, it's also the end of the region if ($i == $#data){ push @url_end,$i+1; } } }else{ if ($in_url_region){ ## This is the end of the region push @url_end, $i; $in_url_region=0; } } } $opt_verbose && print STDERR "."; ## Encapsulate all url blocks in {tabular} environments my($this_begin, $this_end, @beg_text, @end_text); @beg_text = (); @end_text = (); while ($#url_begin >= 0){ # Figure out which type of marker we're adding $this_begin=pop @url_begin; $this_end=pop @url_end; if($this_begin>$this_end){ ## Do Beginning first if ($data[$this_begin] =~ m/etxLink/){ @data = (@data[0 .. $this_begin-1],'\noindent \begin{linkblock}', @data[$this_begin .. $#data]); ## Do End @data = (@data[0 .. $this_end-1], '\end{linkblock}', @data[$this_end .. $#data]); }elsif($data[$this_begin] =~ m/etxUrl/){ @data = (@data[0 .. $this_begin-1],'\noindent \begin{urlblock}', @data[$this_begin .. $#data]); ## Do End @data = (@data[0 .. $this_end-1], '\end{urlblock}', @data[$this_end .. $#data]); } # print "begin: $data[$this_begin]\n"; # if ($data[$this_begin-1] =~ m/\S/){ # $data[$this_begin-1] = "$data[$this_begin-1] \\\\\*"; # @data = (@data[0 .. $this_begin-1],'\noindent \begin{urlblock}', @data[$this_begin .. $#data]); # }else{ # @data = (@data[0 .. $this_begin-1], '\noindent \begin{urlblock}', @data[$this_begin .. $#data]); # } # ## Do End # @data = (@data[0 .. $this_end-1], '\end{urlblock}', @data[$this_end .. $#data]); }else{ # ## Do End first # @data = (@data[0 .. $this_end-1], '\end{urlblock}', @data[$this_end .. $#data]); if ($data[$this_begin] =~ m/etxLink/){ ## Do End first @data = (@data[0 .. $this_end-1], '\end{linkblock}', @data[$this_end .. $#data]); ## Do Beginning @data = (@data[0 .. $this_begin-1],'\noindent \begin{linkblock}', @data[$this_begin .. $#data]); }elsif ($data[$this_begin] =~ m/etxUrl/){ ## Do End first @data = (@data[0 .. $this_end-1], '\end{urlblock}', @data[$this_end .. $#data]); ## Do Beginning @data = (@data[0 .. $this_begin-1],'\noindent \begin{urlblock}', @data[$this_begin .. $#data]); } # if ($data[$this_begin-1] =~ m/\S/){ # $data[$this_begin-1] = "$data[$this_begin-1] \\\\\*"; # @data = (@data[0 .. $this_begin-1], '\noindent \begin{urlblock}', @data[$this_begin .. $#data]); # }else{ # @data = (@data[0 .. $this_begin-1], '\noindent \begin{urlblock}', @data[$this_begin .. $#data]); # } } } $opt_verbose && print STDERR ".\n"; } ################################################################### # # doItemization() # # Modifies: @data # # Converts paragraph-starting astrisks into bullet marks and groups # sequential bulleted lines into {briefs} environments # sub doItemization{ $opt_verbose && print STDERR "doItemization"; for ($i = 0; $i <= $#data; $i++) { ## Convert ^* to \item $data[$i] =~ s/^\*(\s)/\\item$1/; } # Walk through doc finding itemize regions my(@item_begin, @item_end, $in_item_region); $in_item_region = 0; # (i.e., not in an itemize region) for ($i = 0; $i <= $#data; $i++) { $_ = $data[$i]; # $_ is default for m// if (m/^\\item/ && ! $in_item_region){ ## This is the beginning of a region push @item_begin, $i; $in_item_region=1; } if ($in_item_region && (m/^\\(sub)*section/ || m/^(\W)*\-+(\W)*$/ || m/^(\W)*\\etxBold\{.*\}\s+--/)){ ## This is the end of a region push @item_end, $i; $in_item_region=0; } } # Catch itemize regions that terminate with the end of the file if ($#item_begin > $#item_end){ push @item_end, $#data;} # Alternately could just see if ($in_item_region==1) if ($#item_begin != $#item_end){ print STDERR "begins: @item_begin\n"; print STDERR "ends: @item_end\n"; die("unmatched \\begin{briefs}"); } #Add \begin and \end{briefs} to mark the regions we found while ($#item_begin >= 0){ # Figure out which type of marker we're adding $this_begin=pop @item_begin; $this_end=pop @item_end; if($this_begin>$this_end){ print STDERR "\n$data[$this_begin]\nSTART ITEMIZE\n$data[$this_begin+1]\n"; #@data = (@data[0 .. $this_begin-1], '\begin{briefs}\setlength{\parskip}{0.8ex plus0.2ex minus0.1ex}', @data[$this_begin .. $#data]); @data = (@data[0 .. $this_begin-1], '\begin{briefs}', @data[$this_begin .. $#data]); @data = (@data[0 .. $this_end-1], '\end{briefs}', @data[$this_end .. $#data]); }else{ @data = (@data[0 .. $this_end-1], '\end{briefs}', @data[$this_end .. $#data]); #@data = (@data[0 .. $this_begin-1], '\begin{briefs}\setlength{\parskip}{0.8ex plus0.2ex minus0.1ex}', @data[$this_begin .. $#data]); @data = (@data[0 .. $this_begin-1], '\begin{briefs}', @data[$this_begin .. $#data]); } } $opt_verbose && print STDERR ".\n"; } ################################################################### # # doCleanup() # # Modifies: @data # # Performs a few miscellaneous conversions: # o Smart quotes # o Converting -------'s to \hrule's # o Substituting copyright symbol for (c) # o Placing exponents (e.g., FutureBasic^3) within a math/superscript environment # o Constraining spaces after titles (Mr. Mrs. etc.) not to be inter-sentence spaces # o Unsmallcapify italic smallcaps (which don't exist under latex) # o Makes boilerplate text at end of MDJ issue/after double-dollar really tiny # sub doCleanup{ $opt_verbose && print STDERR "doCleanup"; for ($i = 0; $i <= $#data; $i++) { $_ = $data[$i]; # $_ is default for m// # As before, skip lines that are just URLs next if ( m/^\s*\[(\d+)\]\s+\<([[:print:]]+?)\>\s*$/ || m/^\s*\<([[:print:]]+?)\>\s*/ ); ## Educate single quotes s/(\s)\'/$1\`/g; s/\\,\'/\\,\`/g; s/([\(\[])\'/$1\`/g; s/^\'/\`/g; ## Educate nested single-double/double-single quotes s/\'\"/\'\\,\'\'/g; s/\"\'/\`\`\\,\`/g; ## Educate double quotes s/(\s)\"/$1\`\`/g; # [whitespace]["] s/\\,\"/\\,\`\`/g; # [smallspace]["] s/([\(\[\{])\"/$1\`\`/g; # [opening paren/bracket/brace]["] s/^\"/\`\`/g; # [newline]["] s/\"/\'\'/g; # (remaining quotes should all be closers) ## Substitute \hrules for ---- strings that aren't sections s/^(\s)*\-\-\-(\-)*(\s)*$/\\vspace\{2ex\}\\hrule\\vspace\{2ex\}/; ## Replace (c) with copyright symbol s/\([Cc]\)/\\textcopyright /g; # Render the `disclaimer' symbol differently s/\\etxBold\{\[D\]\}/\\etxBold\{\\textsc\{\[d\]\}\}/g; $data[$i] = $_; # Retain our changes ## Handle exponents if ($isPublic){ $data[$i] =~ s/(\^)(\d+)(W*)/\$$1\{$2\}\$$3/g; }else{ if ($data[$i] =~ m/etxLink\{/){ $data[$i] =~ s/\^/\^\{\}/g; }else{ $data[$i] =~ s/(\^)(\d+)(W*)/\\superior\{$2\}$3/g; } } $data[$i] =~ s/(\^)(?!\{\d)/\\\^{}/g; # Fix spacing after personal titles... $data[$i] =~ s/(\s)(Mr\. )([A-Z])/$1Mr.~$3/; $data[$i] =~ s/(\s)(Mrs\. )([A-Z])/$1Mrs.~$3/; $data[$i] =~ s/(\s)(Ms\. )([A-Z])/$1Ms.~$3/; $data[$i] =~ s/(\s)(Dr\. )([A-Z])/$1Dr.~$3/; $data[$i] =~ s/(\s)(Vs\. )(\w)/$1Vs.~$3/; $data[$i] =~ s/(\s)(v\. )(\w)/$1v.~$3/; $data[$i] =~ s/(\s)(Inc\. )([a-z0-9\$])/$1Inc.~$3/; # $data[$i] =~ s/(\s)(vs\. )(\w)/$1vs.~$3/; $data[$i] =~ s/(\s)(vs\.)(\z|\s)/$1$2\\\@$3/g; # ..and dates $data[$i] =~ s/(\s)(Jan\.) ([0-9])/$1$2~$3/; $data[$i] =~ s/(\s)(Feb\.) ([0-9])/$1$2~$3/; $data[$i] =~ s/(\s)(Mar\.) ([0-9])/$1$2~$3/; $data[$i] =~ s/(\s)(Apr\.) ([0-9])/$1$2~$3/; $data[$i] =~ s/(\s)(Jun\.) ([0-9])/$1$2~$3/; $data[$i] =~ s/(\s)(Jul\.) ([0-9])/$1$2~$3/; $data[$i] =~ s/(\s)(Aug\.) ([0-9])/$1$2~$3/; $data[$i] =~ s/(\s)(Sept\.) ([0-9])/$1$2~$3/; $data[$i] =~ s/(\s)(Oct\.) ([0-9])/$1$2~$3/; $data[$i] =~ s/(\s)(Nov\.) ([0-9])/$1$2~$3/; $data[$i] =~ s/(\s)(Dec\.) ([0-9])/$1$2~$3/; # Convert ... to \ldots $data[$i] =~ s/\.\.\.+(\s)/\\ldots\\$1/gi; $data[$i] =~ s/\.\.\.+/\\ldots /gi; # # Correct for italic smallcaps not existing # # $data[$i] =~ s/(\\etxEmph\{\\textsc\{)(.*?)(\}\})/\\etxEmph{\U$2}/g; # $data[$i] =~ s/(\\textsc\{\\etxEmph\{)(.*?)(\}\})/\\etxEmph{\U$2}/g; # -or- # #$nop="[^\{]*?"; # Non-Opeining Paren chars $ncp="[^\}]*?"; # Non-Closing Paren chars if ($data[$i] =~ m/\\textsc/ && ($data[$i] =~ m/\\etxEmph/ || $data[$i] =~ m/\\etxUline/)) { while ($data[$i] =~ m/(\\textsc\{)($ncp)(\})/g){ $pre="$`"; $post="$'"; $thetag="$&"; if ($thetag =~ m/\\etxEmph/ || $thetag =~ m/\\etxUline/){ # catch \textsc{\etxEmph{ $thetag =~ s/\\textsc\{(\\.*?\{)($ncp)\}/$1\U$2/; $data[$i] = "$pre$thetag$post"; }elsif($pre =~ m/\\etxEmph\{$/s || $pre =~ m/\\etxUline\{$/s ){ # catch \etxEmph{\textsc{ $thetag =~ s/\\textsc\{([^\\]$ncp)\}/\U$1/; $data[$i] = "$pre$thetag$post"; }else{ # catch \etxEmph{some other words \textsc{capword}} $lastopen = rindex $pre, '\etxEmph'; if ($lastopen == -1){$lastopen = rindex $pre, '\etxUline';} $lastclose = index $pre, '}'; if ($lastclose==-1 && $lastopen != -1){ $thetag =~ s/\\textsc\{([^\\]$ncp)\}/\U$1/; $data[$i] = "$pre$thetag$post"; } } } } } # Shrink setext bolierplate text (after the $$ end-of-doc tag) if(!$opt_mdj){ for ($i = 0; $i <= $#data; $i++) { if ($data[$i] =~ m/^\\\$\\\$/){ $data[$i] = '\vspace{2ex}\hrule\vspace{2ex} \relsize{-2}'; last; } } } ## Make the MDJ boilerplate more like boilerplate if($opt_mdj){ my $final_hrule=0; for ($i = 0; $i <= $#data; $i++) { $_ = $data[$i]; # $_ is default for m// # find final hrule if (m/\\hrule/){ # If there's more than one `hrule' in an MDJ issue, # it's probably a typo in a Section tag. So complain: if ($final_hrule != 0){ print STDERR "Suspicious hrule:\n"; print STDERR " $data[$final_hrule-2]\n $data[$final_hrule-1]\n $data[$final_hrule]\n\n"; } $final_hrule = $i; } } if ($final_hrule){ # Shrink boilerplate text $data[$final_hrule] = join('',"$data[$final_hrule]",' \\relsize{-2}','\setlength{\parskip}{2ex}','\raggedright'); ################################### # Make the staff listing look nicer ################################### my $staff_begins=-1; my $staff_ends=-1; # Find the block of text for ($i = $final_hrule; $i <= $#data; $i++) { $_ = $data[$i]; if (m/^\s*?Publisher/){ if ($staff_begins != -1){ print STDERR "Suspicious mdj publisher line:\n"; print STDERR " $data[$staff_begins-2]\n $data[$staff_begins-1]\n $data[$staff_begins]\n\n"; } $staff_begins = $i; } if ($staff_begins != -1 and m/^\s*$/){ $staff_ends = $i-1; last; } } # Put the block contents into a tabular if ($staff_begins < $staff_ends and $staff_ends != -1){ @staff_table = @data[$staff_begins..$staff_ends]; for ($i = 0; $i <= $#staff_table; $i++) { $newline = $staff_table[$i]; if ($newline =~ /:/){ $newline =~ s/:/:&/; }else{ $newline = "&$newline"; } $newline =~ s/(\\etxInline.*)$/& $1\\\\/; $staff_table[$i] = $newline; } @staff_table = ('\begin{tabular}{rll}', @staff_table, '\end{tabular}', ''); #@data[$staff_begins..$staff_ends] = @staff_table; @data = (@data[0..$staff_begins-1], @staff_table, @data[$staff_ends+1..$#data]); } } } $opt_verbose && print STDERR ".\n"; } ################################################################### # # packageDocument() # # Calls: getLatexFrontmatter, getLatexDocBegin # Modifies: @data # # Packages up lines in @data by adding the proper laTeX preamble, # definitions, and closing. # sub packageDocument{ $opt_verbose && print STDERR "newPackageDocument"; ## PrepareLaTeX header preamble and command declarations my(@latex_header, @latex_begin, @latex_end); @latex_header = &getLatexFrontmatter; @latex_begin = &getLatexDocBegin; @latex_end=('\end{document}', '% Local Variables:', '% TeX-command-default: "PDFLaTeX"', '% End:'); ################################################################ ## Finish Packaging: my @discardedLinesBeforeTitle = @data[0 .. $title_linenum-1]; @data = @data[$title_linenum .. $#data]; # Assemble bulk of output document @data = (@latex_header, @data, @latex_end); # Find proper beginning point for body of document my $author_linenum=-1, $date_linenum=-1; for ($i = $#data; $i >= 0; $i--) { $_ = $data[$i]; # $_ is default for m// ## Look for \title{} \author{} or \date{} if ( ( m/\\author\{.*\}/ || m/\\etxAuthor\{.*\}/) && $author_linenum==-1 ){ $author_linenum = $i; } if ( m/\\date\{.*\}/ && $date_linenum==-1){ $date_linenum = $i; } if ( m/\\etxTitle\{.*\}/ || m/\\author\{.*\}/ || m/\\etxAuthor\{.*\}/ || m/\\date\{.*\}/ ){ # Find latest point in document where a \maketitle # param is defined. This will be the place to put \begin{document} if ($i>$begin_point){ $begin_point=$i; } } } ## Fill in missing \author and \data my(@title_extras); if ($author_linenum==-1){ push @title_extras, '\author{}';} if ($date_linenum==-1){ push @title_extras, '\date{}';} ## Assemble the various parts of the document @data = (@data[0 .. $begin_point], @title_extras, @latex_begin, @data[$begin_point+1 .. $#data]); $opt_verbose && print STDERR ".\n"; } ################################################################### # # outputDocument() # # Reads: $filename # # Write the assembled text to the output file # # sub outputDocument{ ## Dump latex document to file/stdout if ($filename eq "__StdOut__"){ for ($i = 0; $i <= $#data; $i++) {print "$data[$i]\n";} }elsif (! $filename eq ""){ open(OUTFILE, ">$filename") || die("Can't write to file: $!"); for ($i = 0; $i <= $#data; $i++) {print OUTFILE "$data[$i]\n";} close(OUTFILE); }else{ die("No output file specified. (This should have been caught earlier though.)"); } } ################################################################### # # getLatexFrontmatter() # # Reads: $isPublic, $opt_twoup, $opt_numCols, $opt_nofooter # Returns: @_ contains the lines in the LaTeX preamble # sub getLatexFrontmatter{ my(@etx_defs, @latex_header,@latex_normal_header,@latex_twocol_header,@latex_twoup_header); if($isPublic){ ## For public consumption @latex_normal_header=('\documentclass{article}', '\usepackage[pdfpagemode=None,colorlinks,urlcolor=blue,breaklinks=true]{hyperref}', '\usepackage[T1]{fontenc}', '\usepackage{textcomp}', '\usepackage{times}', '\usepackage{relsize}',''); @latex_twocol_header=('\documentclass[twocolumn]{article}', '\usepackage[pdfpagemode=None,colorlinks,urlcolor=blue,breaklinks=true]{hyperref}', '\usepackage[T1]{fontenc}', '\usepackage{textcomp}', '\usepackage{times}', '\usepackage{relsize}','', '%Refine page layout', '\setlength{\columnsep}{0.43in}', '\setlength{\textheight}{9.4in}', '\setlength{\voffset}{-1.15in}', '\setlength{\textwidth}{7.5in}', '\setlength{\hoffset}{-0.5in}', '\setlength{\footskip}{0.5in}', ''); @latex_twoup_header=('\documentclass[twocolumn,landscape]{article}', '\usepackage[pdfpagemode=None,breaklinks=true]{hyperref}', '\usepackage[T1]{fontenc}', '\usepackage{textcomp}', '\usepackage{times}', '\usepackage{relsize}','', '%Resize text to fit landscape-oriented page', '\setlength{\columnsep}{0.7in}', '\setlength{\textheight}{6.6in}', '\setlength{\voffset}{-0.9in}', '\setlength{\footskip}{35pt}','' ); ## Mess w/ font shortcut names @etx_defs=('%setex.pl-specific font aliases', '\newcommand{\textsfc}[1]{\textsf{\fontseries{m}\selectfont #1}}', '\newcommand{\textsfbc}[1]{\textsf{\fontseries{b}\selectfont #1}}', '\newcommand{\textsb}[1]{\textsf{\fontseries{b}\selectfont #1}}','', '%setex.pl-specific formatting commands', '\newcommand{\etxTitle}[1]{\title{\textsfbc{#1}}}', '\newcommand{\etxAuthor}[1]{\author{\textsc{#1}}}', '\newcommand{\etxRef}[1]{{\smaller \textsfbc{[#1]}}}', '\newcommand{\etxInlineUrl}[2]{\noindent {\textsfc{$<$\href{#1}{#2}$>$}}}', '\newcommand{\etxUrl}[2]{\item \textsfc{$<$\href{#1}{#2}$>$}}', '\newcommand{\etxLink}[3]{\item[#1] \textsfc{$<$\smaller\href{#2}{#3}$>$}}', '\newcommand{\etxUline}[1]{\emph{#1}}', '\newcommand{\etxEmph}[1]{\emph{#1}}', '\newcommand{\etxBold}[1]{\textbf{#1}}','', '\newcommand{\topicpar}[1]{\bigskip\noindent\textsf{\bfseries\larger#1}}', '%setex.pl-specific text environments', '\newenvironment{briefs}{\begin{list}{\textbullet}{\setlength{\topsep}{1.0ex plus0.2ex minus0.1ex}\setlength{\itemsep}{0.2ex plus0.2ex minus0.2ex}\setlength{\parskip}{0.8ex plus0.2ex minus0.1ex}\setlength{\parsep}{0.8ex plus0.2ex minus0.2ex}}}{\end{list}}', '\newenvironment{linkblock}[1][1]{\begin{list}{}{\sffamily\fontseries{mc}\selectfont\setlength{\labelsep}{0em}\setlength{\labelwidth}{1.8em}\setlength{\itemindent}{-0.4em}\setlength{\topsep}{-0.3ex plus0.0ex minus0.1ex}\setlength{\itemsep}{0.0ex}\setlength{\parsep}{0.0ex}}\renewcommand{\makelabel}[1]{\smaller\textsfbc{[##1]}}}{\end{list}\vspace{2ex}}', #'\newenvironment{linkblock}[1][1]{\begin{list}{}{\setlength{\labelsep}{0.8ex}\setlength{\labelwidth}{2ex}\setlength{\topsep}{-0.3ex plus0.0ex minus0.1ex}\setlength{\itemsep}{0.0ex}\setlength{\parsep}{0.0ex}\setlength{\itemindent}{-2.3ex}}\renewcommand{\makelabel}[1]{\smaller\textsfbc{[##1]}}}{\end{list}\vspace{2ex}}', '\newenvironment{urlblock}[1][1]{\begin{list}{}{\sffamily\fontseries{mc}\selectfont\setlength{\labelsep}{0.0em}\setlength{\labelwidth}{0em}\setlength{\topsep}{-0.3ex plus0.0ex minus0.1ex}\setlength{\itemsep}{0.0ex}\setlength{\parsep}{0.0ex}\setlength{\itemindent}{-2em}}}{\end{list}\vspace{2ex}}', #'\newenvironment{urlblock}[1][1]{\begin{list}{}{\setlength{\labelsep}{0.0ex}\setlength{\labelwidth}{0ex}\setlength{\topsep}{-0.3ex plus0.0ex minus0.1ex}\setlength{\itemsep}{0.0ex}\setlength{\parsep}{0.0ex}\setlength{\itemindent}{-5.4ex}}}{\end{list}\vspace{2ex}}','', '%catcode to allow `_\' character so URLs will work', '\catcode`\_=12\relax',''); }else{ ## My personal settings @latex_normal_header=('\documentclass[marginkerning,rich,nonums]{article}', '\usepackage{mystyle}', ''); @latex_twocol_header=('\documentclass[marginkerning,twocolumn,wide,rich,nonums]{article}', '\usepackage{mystyle}','', '%Refine page layout', '\setlength{\textheight}{9.4in}', '\setlength{\voffset}{-1.15in}', '\setlength{\textwidth}{7.5in}', '\setlength{\hoffset}{-0.5in}', '\setlength{\footskip}{0.5in}', ''); @latex_twoup_header=('\documentclass[marginkerning,twocolumn,landscape,rich,nonums]{article}', '\usepackage{mystyle}',''); @etx_defs=('%setex.pl-specific formatting commands', '\newcommand{\etxTitle}[1]{\title{\textsfbc{#1}}}', '\newcommand{\etxAuthor}[1]{\author{\textsc{#1}}}', '\newcommand{\etxRef}[1]{{\textsfbc{[#1]}}}', '\newcommand{\etxInlineUrl}[2]{\noindent {\textsfc{<\href{#1}{#2}>}}}', '\newcommand{\etxUrl}[2]{\item \textsfc{<\href{#1}{#2}>}}', '\newcommand{\etxLink}[3]{\item[#1] \textsfc{<\href{#2}{#3}>}}', '\newcommand{\etxUline}[1]{\emph{#1}}', '\newcommand{\etxEmph}[1]{\emph{#1}}', '\newcommand{\etxBold}[1]{\textbf{#1}}','', '\newcommand{\topicpar}[1]{\bigskip\noindent\textsf{\bfseries\larger#1}\\\\[1ex]}', '%setex.pl-specific text environments', '\newenvironment{briefs}{\begin{list}{\textbullet}{\setlength{\topsep}{1.0ex plus0.2ex minus0.1ex}\setlength{\itemsep}{0.2ex plus0.2ex minus0.2ex}\setlength{\parskip}{0.8ex plus0.2ex minus0.1ex}\setlength{\parsep}{0.8ex plus0.2ex minus0.2ex}}}{\end{list}}', '\newenvironment{linkblock}[1][1]{\begin{list}{}{\sffamily\fontseries{mc}\selectfont\setlength{\labelsep}{0em}\setlength{\labelwidth}{1.8em}\setlength{\itemindent}{-0.4em}\setlength{\topsep}{-0.3ex plus0.0ex minus0.1ex}\setlength{\itemsep}{0.0ex}\setlength{\parsep}{0.0ex}}\renewcommand{\makelabel}[1]{\smaller\textsfbc{[##1]}}}{\end{list}\vspace{1ex}}', '\newenvironment{urlblock}[1][1]{\begin{list}{}{\sffamily\fontseries{mc}\selectfont\setlength{\labelsep}{0.0em}\setlength{\labelwidth}{0em}\setlength{\topsep}{-0.3ex plus0.0ex minus0.1ex}\setlength{\itemsep}{0.0ex}\setlength{\parsep}{0.0ex}\setlength{\itemindent}{-2em}}}{\end{list}\vspace{1ex}}' ); } # Choose proper header based on command line args if ($opt_twoup){ @latex_header = @latex_twoup_header; }elsif ($opt_numCols==2){ @latex_header = @latex_twocol_header; }else{ @latex_header = @latex_normal_header; } # Add our custom hyphenations & version boxes if ( 1 ){ my (@hyphenationCommads); @hyphenationCommads=("%Custom hyphenation patterns for words with \`interCapping\'", "\\hyphenation\{@customHyphenations\}",''); push @latex_header, @hyphenationCommads; my (@versionboxes); @versionboxes=('%Define shortcuts for drawing boxed 9s and Xs', '\usepackage{color}', '\setlength{\fboxsep}{0.4ex}', '\setlength{\fboxrule}{1.2pt}', '\newcommand{\ninebox}{{\setlength{\unitlength}{1ex}\begin{picture}(2,2)\thicklines\color[gray]{0.5}\put(-.1,-.1){\framebox(1.645,2){}}\color[rgb]{0.8,0.3,0.1}\put(0.2,0.5){9}\end{picture}}\xspace}', '\newcommand{\tenbox}{{\setlength{\unitlength}{1ex}\begin{picture}(2,2)\thicklines\color[gray]{0.5}\put(-.1,-.1){\framebox(1.645,2){}}\color[rgb]{0.4,0.4,0.6}\put(0.1,0.1){\textsfb{X}}\end{picture}}\xspace}', '\newcommand{\versionNineBox}{\ninebox}', '\newcommand{\versionTenBox}{\tenbox}', '\newcommand{\versionNineAndTenBox}{\ninebox \tenbox}', '\renewcommand{\versionNineBox}{\raisebox{0.2ex}{{\relsize{-3}\parbox{1.4em}{\color[gray]{0.5}\fbox{\textsfbc{9}}}}}}', '\renewcommand{\versionTenBox}{\raisebox{0.2ex}{{\relsize{-3}\raisebox{-0.01ex}{\parbox{1.4em}{\color[gray]{0.5}\fbox{\textsfbc{X\parbox[b][1.48ex]{0em}{}}}}}}}}', '\renewcommand{\versionNineAndTenBox}{\versionNineBox{}\versionTenBox{}}', ); push @latex_header, @versionboxes; } # Add footer if requested if (! $opt_nofooter){ my (@fancy_footer); @fancy_footer=('%Add a footer with the document title to the bottom of each page', '\usepackage{fancyheadings}', '\pagestyle{fancy}', "\\lfoot{\\textsfc{$title_text}}", '\rfoot{\thepage}', '\setlength{\footrulewidth}{0.4pt}', '\lhead{}\rhead{}\chead{}\cfoot{}', '\setlength{\headrulewidth}{0pt}',''); push @latex_header, @fancy_footer; } push @latex_header, @etx_defs; return @latex_header; } ################################################################### # # getLatexDocBegin() # # Reads: $isPublic, $opt_nofooter # Returns: @_ contains the lines beginning the document body # sub getLatexDocBegin{ my(@latex_begin); ## Unchanged fragments @latex_begin=('','\begin{document}', '\raggedbottom', '\small', '\maketitle'); if (! $opt_nofooter){ # Draw the complicated footer even on the title page push @latex_begin, '\thispagestyle{fancy}'; } if (! $isPublic){ # Don't assume that protcode.tex was included in other people's default configs push @latex_begin, '\marginkern{}'; push @latex_begin, '\tableofcontents'; } return @latex_begin; } ################################################################### # # checkLineBreak() # # Calls: getLatexFrontmater # Args: @_ [0]: The type of section (\section, \subsection, \subsubsection) # [1]: The text to be typeset # Returns: 1 if text has a line break in it (given font for section type) # 0 if text fits on a single line # sub checkLineBreak{ my ($title_text, $section_tag, @test_doc, $return_value, $i); $section_tag= shift; $title_text = shift; # We don't want to presume that other people have their TeX system set up already, # so just count characters instead of actually typesetting the line if ($isPublic){ $line_length = length "$title_text"; if ($opt_twoup || $opt_numCols == 1){ # Wide column if ($line_length >= 43){ return 1; }else{ return 0; } }else{ # Narrow column if ($line_length >= 23){ return 1; }else{ return 0; } } } push @test_doc, (&getLatexFrontmatter, '\usepackage{ifthen}', '\newlength{\\titleWidth}', '\begin{document}'); if ($section_tag =~ m/subsubsection/){ push @test_doc, '\normalfont\normalsize\fontseries{sb}\selectfont'; }elsif($section_tag =~ m/subsection/){ push @test_doc, '\normalfont\large\fontseries{sb}\selectfont'; }elsif($section_tag =~ m/section/){ push @test_doc, '\normalfont\Large\bfseries'; }else{ return 0; } push @test_doc, ("\\settowidth\{\\titleWidth\}\{$title_text\}", '\ifthenelse{\lengthtest{\titleWidth > \columnwidth}}{\typeout{SeTeX_OVERFLOW_ERROR}}{}', "$title_text", '\end{document}'); $tempdir = `tempfile -d /tmp/`; chomp $tempdir; system("rm $tempdir && mkdir $tempdir"); open(MYTEST, ">$tempdir/wtest.tex"); for ($i=0; $i<=$#test_doc; $i++){ print MYTEST "$test_doc[$i]\n"; } close(MYTEST); open (TESTOUTPUT, "cd $tempdir && pdflatex $tempdir/wtest |"); while (){ if (m/SeTeX_OVERFLOW_ERROR/g){ $return_value=1; } } system("rm -rf $tempdir"); return $return_value; }