--- loncom/homework/cleanxml/post_xml.pm 2015/12/03 20:40:31 1.1 +++ loncom/homework/cleanxml/post_xml.pm 2015/12/15 15:33:30 1.3 @@ -1,7 +1,7 @@ # The LearningOnline Network # Third step to clean a file. # -# $Id: post_xml.pm,v 1.1 2015/12/03 20:40:31 damieng Exp $ +# $Id: post_xml.pm,v 1.3 2015/12/15 15:33:30 damieng Exp $ # # Copyright Michigan State University Board of Trustees # @@ -42,8 +42,6 @@ use XML::LibXML; use HTML::TokeParser; # used to parse sty files use Tie::IxHash; # for ordered hashes -use Env qw(RES_DIR); # path of res directory parent (without the / at the end) - no warnings 'recursion'; # yes, fix_paragraph is using heavy recursion, I know # these are constants @@ -77,9 +75,13 @@ my $warnings; # 1 = print warnings # Parses the XML document and fixes many things to turn it into a document matching the schema. -# Returns the text of the document as a byte string. +# @param {reference} textref - reference to the text of the document +# @param {string} file_path - path of the file being processed (we only extract the directory path from it) +# @param {string} res_dir - path of res directory parent (without the / at the end) +# @param {boolean} warn - 1 to print warnings, 0 otherwise +# @returns the text of the document as a byte string. sub post_xml { - my ($textref, $file_path, $warn) = @_; + my ($textref, $file_path, $res_dir, $warn) = @_; $warnings = $warn; my $dom_doc = XML::LibXML->load_xml(string => $textref); @@ -92,10 +94,10 @@ sub post_xml { fix_attribute_case($root); - my $fix_by_hand = replace_m($root); + replace_m($root); my @all_block = (@block_elements, @block_html); - add_sty_blocks($file_path, $root, \@all_block); # must come before the subs using @all_block + add_sty_blocks($file_path, $res_dir, $root, \@all_block); # must come before the subs using @all_block fix_block_styles($root, \@all_block); $root->normalize(); @@ -140,9 +142,6 @@ sub post_xml { replace_tm_dtm($root); - if ($fix_by_hand) { - die "The file has been converted but it should be fixed by hand."; - } return $dom_doc->toString(); # byte string ! } @@ -353,11 +352,9 @@ sub fix_attribute_case { # Replaces m by HTML, tm and/or dtm (which will be replaced by later, but they are useful # to know if the element is a block element or not). # m might contain non-math LaTeX, while tm and dtm may only contain math. -# Returns 1 if the file should be fixed by hand, 0 otherwise. sub replace_m { my ($root) = @_; my $doc = $root->ownerDocument; - my $fix_by_hand = 0; # search for variable declarations my @variables = (); my @scripts = $root->getElementsByTagName('script'); @@ -406,7 +403,6 @@ sub replace_m { if ($warnings) { print "WARNING: is used in a script, it should be converted by hand\n"; } - $fix_by_hand = 1; } } } @@ -421,7 +417,6 @@ sub replace_m { if ($warnings) { print "WARNING: m value is not simple text\n"; } - $fix_by_hand = 1; next; } my $text = $m->firstChild->nodeValue; @@ -552,7 +547,6 @@ sub replace_m { $m->parentNode->replaceChild($fragment, $m); } - return $fix_by_hand; } # Returns the HTML equivalent of LaTeX input, using tth @@ -591,8 +585,10 @@ sub html_to_dom { # Use the linked sty files to guess which newly defined elements should be considered blocks. # Also adds to @containing_styles the sty elements that contain styles. # @param {string} fn - the file path (we only extract the directory path from it) +# @param {string} res_dir - path of res directory parent (without the / at the end) +# @param {Element} root - the root element sub add_sty_blocks { - my ($fn, $root, $all_block) = @_; + my ($fn, $res_dir, $root, $all_block) = @_; my $doc = $root->ownerDocument; my @parserlibs = $doc->getElementsByTagName('parserlib'); my @libs = (); @@ -608,7 +604,7 @@ sub add_sty_blocks { my ($name, $path, $suffix) = fileparse($fn); foreach my $sty (@libs) { if (substr($sty, 0, 1) eq '/') { - $sty = $RES_DIR.$sty; + $sty = $res_dir.$sty; } else { $sty = $path.$sty; }