diff options
Diffstat (limited to 'src/doc_reform/doc_reform.d')
| -rwxr-xr-x | src/doc_reform/doc_reform.d | 678 |
1 files changed, 25 insertions, 653 deletions
diff --git a/src/doc_reform/doc_reform.d b/src/doc_reform/doc_reform.d index 550ee28..d48851b 100755 --- a/src/doc_reform/doc_reform.d +++ b/src/doc_reform/doc_reform.d @@ -62,6 +62,8 @@ import std.process; import doc_reform.meta, + doc_reform.meta.metadoc_harvests_authors, + doc_reform.meta.metadoc_harvests_topics, doc_reform.meta.metadoc_summary, doc_reform.meta.metadoc_harvest, doc_reform.meta.metadoc_from_src, @@ -91,26 +93,7 @@ void main(string[] args) { mixin DocReformBiblio; mixin DocReformRgxInitFlags; mixin outputHub; - struct Harvest { - string title = ""; - string[] author_arr = []; - string author = ""; - string author_surname = ""; - string author_surname_fn = ""; - string language = ""; - string language_original = ""; - string uid = ""; - string date_published = ""; - string[] topic_register_arr = []; - string path_html_seg = ""; - string path_html_scroll = ""; - string path_epub = ""; - string url_html_seg = ""; - string url_html_scroll = ""; - string url_epub = ""; - } - Harvest harvested; - Harvest[] harvests; + auto hvst = DocReformHarvest!(); string flag_action; string arg_unrecognized; enum dAM { abstraction, matters } @@ -282,15 +265,24 @@ void main(string[] args) { || opts["harvest-authors"] || opts["harvest-topics"] ) - ? true - : false; + ? true : false; return _is; } bool harvest_authors() { - return opts["harvest-authors"]; + bool _is = ( + opts["harvest"] + || opts["harvest-authors"] + ) + ? true : false; + return _is; } bool harvest_topics() { - return opts["harvest-topics"]; + bool _is = ( + opts["harvest"] + || opts["harvest-topics"] + ) + ? true : false; + return _is; } bool html() { bool _is; @@ -441,6 +433,7 @@ void main(string[] args) { _is = false; } else if (opts["abstraction"] || concordance + || harvest || html || epub || odt @@ -494,6 +487,7 @@ void main(string[] args) { || concordance || source || pod + || harvest || html || epub || odt @@ -682,19 +676,7 @@ void main(string[] args) { DocReformMetaDocSummary!()(doc_abstraction, doc_matters); } if (doc_matters.opt.action.harvest) { - if (doc_matters.opt.action.harvest_authors) { - } - if (doc_matters.opt.action.harvest_topics) { - } - Harvest[] DocReformMetaDocHarvests()( - Harvest harvested, - Harvest[] harvests, - ) { - harvests ~= harvested; - return harvests; - } - harvested = DocReformMetaDocHarvest!()(doc_matters, harvested); - harvests = DocReformMetaDocHarvests!()(harvested, harvests); + hvst.harvests ~= DocReformMetaDocHarvest!()(doc_matters, hvst); } /+ ↓ debugs +/ if (doc_matters.opt.action.debug_do) { @@ -777,19 +759,7 @@ void main(string[] args) { DocReformMetaDocSummary!()(doc_abstraction, doc_matters); } if (doc_matters.opt.action.harvest) { - if (doc_matters.opt.action.harvest_authors) { - } - if (doc_matters.opt.action.harvest_topics) { - } - Harvest[] DocReformMetaDocHarvests()( - Harvest harvested, - Harvest[] harvests, - ) { - harvests ~= harvested; - return harvests; - } - harvested = DocReformMetaDocHarvest!()(doc_matters, harvested); - harvests = DocReformMetaDocHarvests!()(harvested, harvests); + hvst.harvests ~= DocReformMetaDocHarvest!()(doc_matters, hvst); } /+ ↓ debugs +/ if (doc_matters.opt.action.debug_do) { @@ -827,610 +797,12 @@ void main(string[] args) { } } } - if (_opt_action.very_verbose - && harvests.length > 0 - ) { - auto min_repeat_number = 42; - string[] _document_topic_register; - string[] _topic_register; - string[] _sub_topic_register; - Harvest[][string][string][string][string] subject_trees; - string[] topics = []; - string _auth = ""; - foreach(k, doc_harvest; harvests) { - _topic_register = []; - foreach(topic; doc_harvest.topic_register_arr.sort) { - _sub_topic_register = []; - string _spaces; - string[] subject_tree = topic.split(mkup.sep); - switch (subject_tree.length) { - case 1: - subject_trees[subject_tree[0]]["_a"]["_a"]["_a"] ~= doc_harvest; - break; - case 2: - subject_trees[subject_tree[0]][subject_tree[1]]["_a"]["_a"] ~= doc_harvest; - break; - case 3: - subject_trees[subject_tree[0]][subject_tree[1]][subject_tree[2]]["_a"] ~= doc_harvest; - break; - case 4: - subject_trees[subject_tree[0]][subject_tree[1]][subject_tree[2]][subject_tree[3]] ~= doc_harvest; - break; - default: - break; - } - _topic_register ~= _sub_topic_register.join("\n"); - } - auto char_repeat_number = (doc_harvest.title.length - + doc_harvest.author.length + 16); - char_repeat_number = (char_repeat_number > min_repeat_number) - ? char_repeat_number - : min_repeat_number; - _document_topic_register ~= format( - "\"%s\", %s%s\n%s", - doc_harvest.title, - doc_harvest.author, - (doc_harvest.date_published.length > 0) ? " (" ~ doc_harvest.date_published ~ ")" : "", - _topic_register.sort!("toUpper(a) < toUpper(b)", SwapStrategy.unstable).release.join("\n"), - ); - } - topics ~= format(q"┃<!DOCTYPE html> -<html> -<head> -<meta charset="utf-8"> -<title>Metadata Harvest - Topics</title> -<meta http-equiv="Content-Type" content="text/html;charset=utf-8" /> -<meta name="dc.title" content= "metadata harvest, Topics - information Structuring Universe, Structured information Serialised Units" /> -<meta name="dc.subject" content= "document structuring, ebook, publishing, PDF, LaTeX, XML, ODF, SQL, postgresql, sqlite, electronic book, electronic publishing, electronic document, electronic citation, data structure, citation systems, granular search, digital library" /> -<meta name="generator" content="doc_reform" /> -<link rel="generator" href="http://sisudoc.org" /> -<link href="../../_sisu/css/harvest.css" rel="stylesheet"> -<style TYPE="text/css"> -/* DocReform harvest css default stylesheet */ - body { - color: black; - background: #ffffff; - background-color: #ffffff; - } - a:link { - color: #003399; - text-decoration: none; - } - a:visited { - color: #003399; - text-decoration: none; - } - a:hover { - color: #000000; - background-color: #f9f9aa; - } - a:hover img { - background-color: #ffffff; - } - a:active { - color: #003399; - text-decoration: underline; - } - - .norm, .bold { - line-height: 150%%; - margin-left: 1em; - margin-right: 2em; - margin-top: 10px; - margin-bottom: 0px; - text-indent: 0mm; - } - p, h0, h1, h2, h3, h4, h5, h6, h7 { - display: block; - font-family: verdana, arial, georgia, tahoma, sans-serif, helvetica, times, roman; - font-size: 100%%; - font-weight: normal; - line-height: 150%%; - /* text-align: justify; */ - margin-left: 1em; - text-indent: 0mm; - margin-top: 2px; - margin-bottom: 2px; - margin-right: 6px; - text-align: left; - } - h1 { - font-size: 120%%; - font-weight: bold; - color: white; - background: #000088; - margin-left: 0em; - } - p.work { - font-size: 80%%; - margin-left: 5em; - margin-top: 0px; - margin-bottom: 0px; - margin-right: 6px; - text-align: left; - } - p.author { - font-size: 100%%; - margin-left: 2em; - margin-top: 0px; - margin-bottom: 0px; - margin-right: 6px; - text-align: left; - } - p.publication { - font-size: 80%%; - margin-left: 4em; - margin-top: 0px; - margin-bottom: 0px; - margin-right: 6px; - text-align: left; - } - p.letter { - font-weight: bold; - font-size: 60%%; - margin-left: 1em; - margin-top: 0px; - margin-bottom: 0px; - margin-right: 6px; - text-align: left; - color: white; - background: #880000; - } - p.lev0 { - font-size: 120%%; - margin-left: 1em; - color: white; - background: #000000; - } - - p.lev1 { - font-size: 110%%; - margin-left: 2em; - color: white; - background: #444444; - } - p.lev2 { - font-size: 100%%; - margin-left: 3em; - background: #888888; - } - p.lev3 { - font-size: 90%%; - margin-left: 4em; - background: #bbbbbb; - } - p.lev4 { - font-size: 80%%; - margin-left: 5em; - background: #eeeeee; - } - p.lev5 { - font-size: 80%%; - margin-left: 6em; - } -</style> -<link rel="shortcut icon" href="../_sisu/image/rb7.ico" /> -</head> -<body lang="en" xml:lang="en"> -<a name="top" id="top"></a> -<a name="up" id="up"></a> -<a name="start" id="start"></a> -<h1>Metadata Harvest - Topics (output organised by language & filetype)</h1> -<p>[<a href="../../index.html"> HOME </a>] also see <a href="authors.html">Metadata Harvest - Authors</a></p> -<p><a href="#A">A</a>, <a href="#B">B</a>, <a href="#C">C</a>, <a href="#D">D</a>, <a href="#E">E</a>, <a href="#F">F</a>, <a href="#G">G</a>, <a href="#H">H</a>, <a href="#I">I</a>, <a href="#J">J</a>, <a href="#K">K</a>, <a href="#L">L</a>, <a href="#M">M</a>, <a href="#N">N</a>, <a href="#O">O</a>, <a href="#P">P</a>, <a href="#Q">Q</a>, <a href="#R">R</a>, <a href="#S">S</a>, <a href="#T">T</a>, <a href="#U">U</a>, <a href="#V">V</a>, <a href="#W">W</a>, <a href="#X">X</a>, <a href="#Y">Y</a>, <a href="#Z">Z</a>, -<p></p> -<hr /> -<p class="tiny"><a href="../../en/manifest/topics.html">English</a> </p> -<hr /> -┃") ~ "\n"; - char _prev_k = "_".to!char; - int _kn; - foreach(k0; - subject_trees.keys - .sort!("toUpper(a) < toUpper(b)", SwapStrategy.unstable) - ) { - if (k0.toUpper.to!(char[])[0] != _prev_k) { - topics ~= format(q"┃<p class="letter"><a name="%s">%s</a></p><p class="book_index_lev1"><a name="a"></a></p>┃", - k0.toUpper.to!(char[])[0], - k0.toUpper.to!(char[])[0], - ); - _prev_k = k0.toUpper.to!(char[])[0]; - } - if (k0 != "_a") { - topics ~= format(q"┃<p class="lev0"><a name="%s">%s</a></p>┃", - k0, k0,) ~ "\n"; - writeln("", k0); - if ("_a" in subject_trees[k0]) { - foreach (t_a_; - subject_trees[k0]["_a"]["_a"]["_a"] - .multiSort!("toUpper(a.title) < toUpper(b.title)", "a.author < b.author", SwapStrategy.unstable) - ) { - _auth = []; - if (t_a_.author_arr.length < 2) { - _auth = format(q"┃ <a href="authors.html#%s">%s</a>┃", - t_a_.author_surname, - t_a_.author, - ); - } else { - foreach (a; t_a_.author_arr) { - _auth ~= format(q"┃ <a href="authors.html#%s">%s</a>,┃", - t_a_.author_surname, - a, - ); - } - } - topics ~= format(q"┃<p class="work"><a href="%s">"%s"</a> -%s┃", - "url", - t_a_.title, - _auth, - ) ~ "\n"; - writeln("- ", t_a_.title, " - ", t_a_.author); - } - } - foreach(k1; - subject_trees[k0].keys - .sort!("toUpper(a) < toUpper(b)", SwapStrategy.unstable) - ) { - if (k1 != "_a") { - topics ~= format(q"┃<p class="lev1"><a name="%s">%s</a></p>┃", - k1, k1,) ~ "\n"; - writeln(" ", k1); - if ("_a" in subject_trees[k0][k1]) { - foreach (t_a_; - subject_trees[k0][k1]["_a"]["_a"] - .multiSort!("toUpper(a.title) < toUpper(b.title)", "a.author < b.author", SwapStrategy.unstable) - ) { - _auth = []; - if (t_a_.author_arr.length < 2) { - _auth = format(q"┃ <a href="authors.html#%s">%s</a>┃", - t_a_.author_surname, - t_a_.author, - ); - } else { - foreach (a; t_a_.author_arr) { - _auth ~= format(q"┃ <a href="authors.html#%s">%s</a>,┃", - t_a_.author_surname, - a, - ); - } - } - topics ~= format(q"┃<p class="work"><a href="%s">%s</a> -%s┃", - "url", - t_a_.title, - _auth, - ) ~ "\n"; - writeln(" - ", t_a_.title, " - ", t_a_.author); - } - } - } - foreach(k2; - subject_trees[k0][k1].keys - .sort!("toUpper(a) < toUpper(b)", SwapStrategy.unstable) - ) { - if (k2 != "_a") { - topics ~= format(q"┃<p class="lev2"><a name="%s">%s</a></p>┃", - k2, k2,) ~ "\n"; - writeln(" ", k2); - if ("_a" in subject_trees[k0][k1][k2]) { - foreach (t_a_; - subject_trees[k0][k1][k2]["_a"] - .multiSort!("toUpper(a.title) < toUpper(b.title)", "a.author < b.author", SwapStrategy.unstable) - ) { - _auth = []; - if (t_a_.author_arr.length < 2) { - _auth = format(q"┃ <a href="authors.html#%s">%s</a>┃", - t_a_.author_surname, - t_a_.author, - ); - } else { - foreach (a; t_a_.author_arr) { - _auth ~= format(q"┃ <a href="authors.html#%s">%s</a>,┃", - t_a_.author_surname, - a, - ); - } - } - topics ~= format(q"┃<p class="work"><a href="%s">%s</a> -%s┃", - "url", - t_a_.title, - _auth, - ) ~ "\n"; - writeln(" - ", t_a_.title, " - ", t_a_.author); - } - } - } - foreach(k3; - subject_trees[k0][k1][k2].keys - .sort!("toUpper(a) < toUpper(b)", SwapStrategy.unstable) - ) { - if (k3 != "_a") { - topics ~= format(q"┃<p class="lev3"><a name="%s">%s</a></p>┃", - k3, k3,) ~ "\n"; - writeln(" ", k3); - { - foreach (t_a_; - subject_trees[k0][k1][k2][k3] - .multiSort!("toUpper(a.title) < toUpper(b.title)", "a.author < b.author", SwapStrategy.unstable) - ) { - _auth = []; - if (t_a_.author_arr.length < 2) { - _auth = format(q"┃<a href="authors.html#%s">%s</a>┃", - t_a_.author_surname, - t_a_.author, - ); - } else { - foreach (a; t_a_.author_arr) { - _auth ~= format(q"┃ <a href="authors.html#%s">%s</a>,┃", - t_a_.author_surname, - a, - ); - } - } - topics ~= format(q"┃ <p class="work"><a href="%s">%s</a> -%s┃", - "url", - t_a_.title, - _auth, - ) ~ "\n"; - writeln(" - ", t_a_.title, " - ", t_a_.author); - } - } - } - } - } - } - } - } - topics ~= format(q"┃ -<hr /> -<a name="bottom" id="bottom"></a> -<a name="down" id="down"></a> -<a name="end" id="end"></a> -<a name="finish" id="finish"></a> -<a name="stop" id="stop"></a> -<a name="credits"></a> -</body> -</html> -┃") ~ "\n"; - try { - auto f = File("topics.html", "w"); - foreach (o; topics) { - f.writeln(o); - } - } catch (ErrnoException ex) { - // Handle error - } - } - if ((_opt_action.verbose - || _opt_action.very_verbose) - && harvests.length > 0 - ) { - string[] authors = []; - authors ~= format(q"┃ -<!DOCTYPE html> -<html> -<head> -<meta charset="utf-8"> -<title>Metadata Harvest - Authors</title> -<meta http-equiv="Content-Type" content="text/html;charset=utf-8" /> -<meta name="dc.title" content= "metadata harvest, Authors - information Structuring Universe, Structured information Serialised Units" /> -<meta name="dc.subject" content= "document structuring, ebook, publishing, PDF, LaTeX, XML, ODF, SQL, postgresql, sqlite, electronic book, electronic publishing, electronic document, electronic citation, data structure, citation systems, granular search, digital library" /> -<meta name="generator" content="doc_reform" /> -<link rel="generator" href="http://sisudoc.org" /> -<link href="../../_sisu/css/harvest.css" rel="stylesheet" > -<style TYPE="text/css"> -/* DocReform harvest css default stylesheet */ - body { - color: black; - background: #ffffff; - background-color: #ffffff; - } - a:link { - color: #003399; - text-decoration: none; - } - a:visited { - color: #003399; - text-decoration: none; - } - a:hover { - color: #000000; - background-color: #f9f9aa; - } - a:hover img { - background-color: #ffffff; - } - a:active { - color: #003399; - text-decoration: underline; - } - - .norm, .bold { - line-height: 150%%; - margin-left: 1em; - margin-right: 2em; - margin-top: 10px; - margin-bottom: 0px; - text-indent: 0mm; - } - p, h0, h1, h2, h3, h4, h5, h6, h7 { - display: block; - font-family: verdana, arial, georgia, tahoma, sans-serif, helvetica, times, roman; - font-size: 100%%; - font-weight: normal; - line-height: 150%%; - /* text-align: justify; */ - margin-left: 1em; - text-indent: 0mm; - margin-top: 2px; - margin-bottom: 2px; - margin-right: 6px; - text-align: left; - } - h1 { - font-size: 120%%; - font-weight: bold; - color: white; - background: #000088; - margin-left: 0em; - } - p.work { - font-size: 80%%; - margin-left: 5em; - margin-top: 0px; - margin-bottom: 0px; - margin-right: 6px; - text-align: left; - } - p.author { - font-size: 100%%; - margin-left: 2em; - margin-top: 0px; - margin-bottom: 0px; - margin-right: 6px; - text-align: left; - } - p.publication { - font-size: 80%%; - margin-left: 4em; - margin-top: 0px; - margin-bottom: 0px; - margin-right: 6px; - text-align: left; - } - p.letter { - font-weight: bold; - font-size: 60%%; - margin-left: 1em; - margin-top: 0px; - margin-bottom: 0px; - margin-right: 6px; - text-align: left; - color: white; - background: #880000; - } - p.lev0 { - font-size: 120%%; - margin-left: 1em; - color: white; - background: #000000; - } - - p.lev1 { - font-size: 110%%; - margin-left: 2em; - color: white; - background: #444444; - } - p.lev2 { - font-size: 100%%; - margin-left: 3em; - background: #888888; - } - p.lev3 { - font-size: 90%%; - margin-left: 4em; - background: #bbbbbb; - } - p.lev4 { - font-size: 80%%; - margin-left: 5em; - background: #eeeeee; - } - p.lev5 { - font-size: 80%%; - margin-left: 6em; - } -</style> -<link rel="shortcut icon" href="../_sisu/image/rb7.ico" /> -</head> -<body lang="en" xml:lang="en"> -<a name="top" id="top"></a> -<a name="up" id="up"></a> -<a name="start" id="start"></a> -<h1>Metadata Harvest - Authors (output organised by language & filetype)</h1> -<p>[<a href="../../index.html"> HOME </a>] also see <a href="topics.html">Metadata Harvest - Topics</a></p> -<p></p> -<hr /> -<p class="tiny"><a href="../../en/manifest/authors.html">English</a> </p> -<hr /> -<p><a href="#A">A</a>, <a href="#B">B</a>, <a href="#C">C</a>, <a href="#D">D</a>, <a href="#E">E</a>, <a href="#F">F</a>, <a href="#G">G</a>, <a href="#H">H</a>, <a href="#I">I</a>, <a href="#J">J</a>, <a href="#K">K</a>, <a href="#L">L</a>, <a href="#M">M</a>, <a href="#N">N</a>, <a href="#O">O</a>, <a href="#P">P</a>, <a href="#Q">Q</a>, <a href="#R">R</a>, <a href="#S">S</a>, <a href="#T">T</a>, <a href="#U">U</a>, <a href="#V">V</a>, <a href="#W">W</a>, <a href="#X">X</a>, <a href="#Y">Y</a>, <a href="#Z">Z</a>, -┃") ~ "\n"; - string[string] _au; - string[] _auth_date_title; - string[] _author_date_title; - string _prev_auth = ""; - char _prev_k = "_".to!char; - foreach(doc_harvest; - harvests - .multiSort!( - "toUpper(a.author_surname_fn) < toUpper(b.author_surname_fn)", - "a.date_published < b.date_published", - "a.title < b.title", - SwapStrategy.unstable - ) - ) { - if (doc_harvest.author_surname_fn != _prev_auth) { - _au[doc_harvest.author_surname_fn] - = format(q"┃<p class="author"><a name="%s">%s</a></p> <p class="publication">%s "<a href="%s">%s</a>" [%s]</p>┃", - doc_harvest.author_surname, - doc_harvest.author_surname_fn, - (doc_harvest.date_published.length > 0) - ? doc_harvest.date_published : "", - "url", - doc_harvest.title, - doc_harvest.language, - ); - _prev_auth = doc_harvest.author_surname_fn; - } else { - _au[doc_harvest.author_surname_fn] - ~= format(q"┃<p class="publication">%s "<a href="%s">%s</a>" [%s]</p>┃", - (doc_harvest.date_published.length > 0) - ? doc_harvest.date_published : "", - "url", - doc_harvest.title, - doc_harvest.language, - ); - } - _author_date_title ~= format(q"┃%s %s "%s" [%s]┃", - doc_harvest.author_surname_fn, - (doc_harvest.date_published.length > 0) - ? "(" ~ doc_harvest.date_published ~ ")" : "", - doc_harvest.title, - doc_harvest.language, - ); - // writeln(doc_harvest.author_date_title); - } - foreach (k; _au.keys.sort) { - if (k.toUpper.to!(char[])[0] != _prev_k) { - authors ~= format(q"┃<p class="letter"><a name="%s">%s</a></p><p class="book_index_lev1"><a name="a"></a></p>┃", - k.toUpper.to!(char[])[0], - k.toUpper.to!(char[])[0], - ); - _prev_k = k.toUpper.to!(char[])[0]; - } - authors ~= _au[k]; - } - authors ~= format(q"┃ -<hr /> -<a name="bottom" id="bottom"></a> -<a name="down" id="down"></a> -<a name="end" id="end"></a> -<a name="finish" id="finish"></a> -<a name="stop" id="stop"></a> -<a name="credits"></a> -</body> -</html> -┃") ~ "\n"; - try { - auto f = File("authors.html", "w"); - foreach (o; authors) { - f.writeln(o); - } - } catch (ErrnoException ex) { - // Handle error + if (hvst.harvests.length > 0) { + if (_opt_action.harvest_topics) { + DocReformMetaDocHarvestsTopics!()(hvst, _opt_action); } - foreach(_adt; _author_date_title.sort) { - writeln(_adt); + if (_opt_action.harvest_authors) { + DocReformMetaDocHarvestsAuthors!()(hvst.harvests, _opt_action); } } } |