]> git.lyx.org Git - features.git/blobdiff - lib/lyx2lyx/lyx_2_3.py
Support for multiple bibliographies
[features.git] / lib / lyx2lyx / lyx_2_3.py
index e3eab80c16fffd74e52b2270fdfa19533ba405bb..3b46470b18dd85af4df9e3d929ccbd4131e40f8a 100644 (file)
@@ -1018,7 +1018,7 @@ def revert_cjkquotes(document):
                 if cjk:
                     replace = [u"\u300A"]
                 else:
-                    replace = ["\\begin_inset Formula $\\langle\\kern -2.5pt\\langle$$", "\\end_inset"]
+                    replace = ["\\begin_inset Formula $\\langle\\kern -2.5pt\\langle$", "\\end_inset"]
             else:
                 # outer closing mark
                 if cjk:
@@ -1082,7 +1082,7 @@ def revert_labelonly(document):
             i = j + 1
             continue
         document.body[i:j+1] = put_cmd_in_ert([label])
-        i = j + 1
+        i += 1
 
 
 def revert_plural_refs(document):
@@ -1101,18 +1101,24 @@ def revert_plural_refs(document):
             i += 1
             continue
 
-        plural = caps = label = False
-        if use_refstyle:
+        plural = caps = suffix = False
+        k = find_token(document.body, "LaTeXCommand formatted", i, j)
+        if k != -1 and use_refstyle:
             plural = get_bool_value(document.body, "plural", i, j, False)
             caps   = get_bool_value(document.body, "caps", i, j, False)
             label  = get_quoted_value(document.body, "reference", i, j)
             if label:
-                (prefix, suffix) = label.split(":", 1)
+                try:
+                    (prefix, suffix) = label.split(":", 1)
+                except:
+                    document.warning("No `:' separator in formatted reference at line %d!" % (i))
             else:
                 document.warning("Can't find label for reference at line %d!" % (i))
 
-        # this effectively tests also for use_refstyle
-        if not ((plural or caps) and label):
+        # this effectively tests also for use_refstyle and a formatted reference
+        # we do this complicated test because we would otherwise do this erasure
+        # over and over and over
+        if not ((plural or caps) and suffix):
             del_token(document.body, "plural", i, j)
             del_token(document.body, "caps", i, j - 1) # since we deleted a line
             i = j - 1
@@ -1125,7 +1131,610 @@ def revert_plural_refs(document):
             cmd += "[s]"
         cmd += "{" + suffix + "}"
         document.body[i:j+1] = put_cmd_in_ert([cmd])
+        i += 1
+
+
+def revert_noprefix(document):
+    " Revert labelonly tags with 'noprefix' set "
+    i = 0
+    while (True):
+        i = find_token(document.body, "\\begin_inset CommandInset ref", i)
+        if i == -1:
+            return
+        j = find_end_of_inset(document.body, i)
+        if j == -1:
+            document.warning("Can't find end of reference inset at line %d!!" %(i))
+            i += 1
+            continue
+        k = find_token(document.body, "LatexCommand labelonly", i, j)
+        if k == -1:
+            i = j
+            continue
+        noprefix = get_bool_value(document.body, "noprefix", i, j)
+        if not noprefix:
+            del_token(document.body, "noprefix", i, j)
+            i = j
+            continue
+        label = get_quoted_value(document.body, "reference", i, j)
+        if not label:
+            document.warning("Can't find label for reference at line %d!" %(i))
+            i = j + 1
+            continue
+        try:
+            (prefix, suffix) = label.split(":", 1)
+        except:
+            document.warning("No `:' separator in formatted reference at line %d!" % (i))
+            # we'll leave this as an ordinary labelonly reference
+            del_token(document.body, "noprefix", i, j)
+            i = j
+            continue
+        document.body[i:j+1] = put_cmd_in_ert([suffix])
+        i += 1
+
+
+def revert_biblatex(document):
+    " Revert biblatex support "
+
+    #
+    # Header
+    #
+
+    # 1. Get cite engine
+    engine = "basic"
+    i = find_token(document.header, "\\cite_engine", 0)
+    if i == -1:
+        document.warning("Malformed document! Missing \\cite_engine")
+    else:
+        engine = get_value(document.header, "\\cite_engine", i)
+
+    # 2. Store biblatex state and revert to natbib
+    biblatex = False
+    if engine in ["biblatex", "biblatex-natbib"]:
+        biblatex = True
+        document.header[i] = "\\cite_engine natbib"
+
+    # 3. Store and remove new document headers
+    bibstyle = ""
+    i = find_token(document.header, "\\biblatex_bibstyle", 0)
+    if i != -1:
+        bibstyle = get_value(document.header, "\\biblatex_bibstyle", i)
+        del document.header[i]
+
+    citestyle = ""
+    i = find_token(document.header, "\\biblatex_citestyle", 0)
+    if i != -1:
+        citestyle = get_value(document.header, "\\biblatex_citestyle", i)
+        del document.header[i]
+
+    biblio_options = ""
+    i = find_token(document.header, "\\biblio_options", 0)
+    if i != -1:
+        biblio_options = get_value(document.header, "\\biblio_options", i)
+        del document.header[i]
+
+    if biblatex:
+        bbxopts = "[natbib=true"
+        if bibstyle != "":
+            bbxopts += ",bibstyle=" + bibstyle
+        if citestyle != "":
+            bbxopts += ",citestyle=" + citestyle
+        if biblio_options != "":
+            bbxopts += "," + biblio_options
+        bbxopts += "]"
+        add_to_preamble(document, "\\usepackage" + bbxopts + "{biblatex}")
+
+    #
+    # Body
+    #
+
+    # 1. Bibtex insets
+    i = 0
+    bibresources = []
+    while (True):
+        i = find_token(document.body, "\\begin_inset CommandInset bibtex", i)
+        if i == -1:
+            break
+        j = find_end_of_inset(document.body, i)
+        if j == -1:
+            document.warning("Can't find end of bibtex inset at line %d!!" %(i))
+            i += 1
+            continue
+        bibs = get_quoted_value(document.body, "bibfiles", i, j)
+        opts = get_quoted_value(document.body, "biblatexopts", i, j)
+        # store resources
+        if bibs:
+            bibresources += bibs.split(",")
+        else:
+            document.warning("Can't find bibfiles for bibtex inset at line %d!" %(i))
+        # remove biblatexopts line
+        k = find_token(document.body, "biblatexopts", i, j)
+        if k != -1:
+            del document.body[k]
+        # Re-find inset end line
+        j = find_end_of_inset(document.body, i)
+        # Insert ERT \\printbibliography and wrap bibtex inset to a Note
+        if biblatex:
+            pcmd = "printbibliography"
+            if opts:
+                pcmd += "[" + opts + "]"
+            repl = ["\\begin_inset ERT", "status open", "", "\\begin_layout Plain Layout",\
+                    "", "", "\\backslash", pcmd, "\\end_layout", "", "\\end_inset", "", "",\
+                    "\\end_layout", "", "\\begin_layout Standard", "\\begin_inset Note Note",\
+                    "status open", "", "\\begin_layout Plain Layout" ]
+            repl += document.body[i:j+1]
+            repl += ["", "\\end_layout", "", "\\end_inset", "", ""]
+            document.body[i:j+1] = repl
+            j += 27
+
         i = j + 1
+
+    if biblatex:
+        for b in bibresources:
+            add_to_preamble(document, "\\addbibresource{" + b + ".bib}")
+
+    # 2. Citation insets
+
+    # Specific citation insets used in biblatex that need to be reverted to ERT
+    new_citations = {
+        "Cite" : "Cite",
+        "citebyear" : "citeyear",
+        "citeyear" : "cite*",
+        "Footcite" : "Smartcite",
+        "footcite" : "smartcite",
+        "Autocite" : "Autocite",
+        "autocite" : "autocite",
+        "citetitle" : "citetitle",
+        "citetitle*" : "citetitle*",
+        "fullcite" : "fullcite",
+        "footfullcite" : "footfullcite",
+        "supercite" : "supercite",
+        "citeauthor" : "citeauthor",
+        "citeauthor*" : "citeauthor*",
+        "Citeauthor" : "Citeauthor",
+        "Citeauthor*" : "Citeauthor*"
+        }
+
+    # All commands accepted by LyX < 2.3. Everything else throws an error.
+    old_citations = [ "cite", "nocite", "citet", "citep", "citealt", "citealp",\
+                     "citeauthor", "citeyear", "citeyearpar", "citet*", "citep*",\
+                      "citealt*", "citealp*", "citeauthor*", "Citet",  "Citep",\
+                      "Citealt",  "Citealp",  "Citeauthor", "Citet*", "Citep*",\
+                      "Citealt*", "Citealp*", "Citeauthor*", "fullcite", "footcite",\
+                      "footcitet", "footcitep", "footcitealt", "footcitealp",\
+                      "footciteauthor", "footciteyear", "footciteyearpar",\
+                     "citefield", "citetitle", "cite*" ]
+
+    i = 0
+    while (True):
+        i = find_token(document.body, "\\begin_inset CommandInset citation", i)
+        if i == -1:
+            break
+        j = find_end_of_inset(document.body, i)
+        if j == -1:
+            document.warning("Can't find end of citation inset at line %d!!" %(i))
+            i += 1
+            continue
+        k = find_token(document.body, "LatexCommand", i, j)
+        if k == -1:
+            document.warning("Can't find LatexCommand for citation inset at line %d!" %(i))
+            i = j + 1
+            continue
+        cmd = get_value(document.body, "LatexCommand", k)
+        if biblatex and cmd in list(new_citations.keys()):
+            pre = get_quoted_value(document.body, "before", i, j)
+            post = get_quoted_value(document.body, "after", i, j)
+            key = get_quoted_value(document.body, "key", i, j)
+            if not key:
+                document.warning("Citation inset at line %d does not have a key!" %(i))
+                key = "???"
+            # Replace known new commands with ERT
+            res = "\\" + new_citations[cmd]
+            if pre:
+                res += "[" + pre + "]"
+            if post:
+                res += "[" + post + "]"
+            elif pre:
+                res += "[]"
+            res += "{" + key + "}"
+            document.body[i:j+1] = put_cmd_in_ert([res])
+        elif cmd not in old_citations:
+            # Reset unknown commands to cite. This is what LyX does as well
+            # (but LyX 2.2 would break on unknown commands)
+            document.body[k] = "LatexCommand cite"
+            document.warning("Reset unknown cite command '%s' with cite" % cmd)
+        i = j + 1
+
+    # Emulate the old biblatex-workaround (pretend natbib in order to use the styles)
+    if biblatex:
+        i = find_token(document.header, "\\begin_local_layout", 0)
+        if i == -1:
+            k = find_token(document.header, "\\language", 0)
+            if k == -1:
+                # this should not happen
+                document.warning("Malformed LyX document! No \\language header found!")
+                return
+            document.header[k-1 : k-1] = ["\\begin_local_layout", "\\end_local_layout"]
+            i = k-1
+
+        j = find_end_of(document.header, i, "\\begin_local_layout", "\\end_local_layout")
+        if j == -1:
+            # this should not happen
+            document.warning("Malformed LyX document! Can't find end of local layout!")
+            return
+
+        document.header[i+1 : i+1] = [
+            "### Inserted by lyx2lyx (biblatex emulation) ###",
+            "Provides natbib 1",
+            "### End of insertion by lyx2lyx (biblatex emulation) ###"
+        ]
+
+
+def revert_citekeyonly(document):
+    " Revert keyonly cite command to ERT "
+
+    i = 0
+    while (True):
+        i = find_token(document.body, "\\begin_inset CommandInset citation", i)
+        if i == -1:
+            break
+        j = find_end_of_inset(document.body, i)
+        if j == -1:
+            document.warning("Can't find end of citation inset at line %d!!" %(i))
+            i += 1
+            continue
+        k = find_token(document.body, "LatexCommand", i, j)
+        if k == -1:
+            document.warning("Can't find LatexCommand for citation inset at line %d!" %(i))
+            i = j + 1
+            continue
+        cmd = get_value(document.body, "LatexCommand", k)
+        if cmd != "keyonly":
+            i = j + 1
+            continue
+
+        key = get_quoted_value(document.body, "key", i, j)
+        if not key:
+            document.warning("Citation inset at line %d does not have a key!" %(i))
+        # Replace known new commands with ERT
+        document.body[i:j+1] = put_cmd_in_ert([key])
+        i = j + 1
+
+
+
+def revert_bibpackopts(document):
+    " Revert support for natbib/jurabib package options "
+
+    engine = "basic"
+    i = find_token(document.header, "\\cite_engine", 0)
+    if i == -1:
+        document.warning("Malformed document! Missing \\cite_engine")
+    else:
+        engine = get_value(document.header, "\\cite_engine", i)
+
+    biblatex = False
+    if engine not in ["natbib", "jurabib"]:
+        return
+
+    i = find_token(document.header, "\\biblio_options", 0)
+    if i == -1:
+        # Nothing to do if we have no options
+        return
+
+    biblio_options = get_value(document.header, "\\biblio_options", i)
+    del document.header[i]
+
+    if not biblio_options:
+        # Nothing to do for empty options
+        return
+
+    i = find_token(document.header, "\\begin_local_layout", 0)
+    if i == -1:
+        k = find_token(document.header, "\\language", 0)
+        if k == -1:
+            # this should not happen
+            document.warning("Malformed LyX document! No \\language header found!")
+            return
+        document.header[k-1 : k-1] = ["\\begin_local_layout", "\\end_local_layout"]
+        i = k - 1
+
+    j = find_end_of(document.header, i, "\\begin_local_layout", "\\end_local_layout")
+    if j == -1:
+        # this should not happen
+        document.warning("Malformed LyX document! Can't find end of local layout!")
+        return
+
+    document.header[i+1 : i+1] = [
+        "### Inserted by lyx2lyx (bibliography package options) ###",
+        "PackageOptions " + engine + " " + biblio_options,
+        "### End of insertion by lyx2lyx (bibliography package options) ###"
+    ]
+
+
+def revert_qualicites(document):
+    " Revert qualified citation list commands to ERT "
+
+    # Citation insets that support qualified lists, with their LaTeX code
+    ql_citations = {
+        "cite" : "cites",
+        "Cite" : "Cites",
+        "citet" : "textcites",
+        "Citet" : "Textcites",
+        "citep" : "parencites",
+        "Citep" : "Parencites",
+        "Footcite" : "Smartcites",
+        "footcite" : "smartcites",
+        "Autocite" : "Autocites",
+        "autocite" : "autocites",
+        }
+
+    # Get cite engine
+    engine = "basic"
+    i = find_token(document.header, "\\cite_engine", 0)
+    if i == -1:
+        document.warning("Malformed document! Missing \\cite_engine")
+    else:
+        engine = get_value(document.header, "\\cite_engine", i)
+
+    biblatex = engine in ["biblatex", "biblatex-natbib"]
+
+    i = 0
+    while (True):
+        i = find_token(document.body, "\\begin_inset CommandInset citation", i)
+        if i == -1:
+            break
+        j = find_end_of_inset(document.body, i)
+        if j == -1:
+            document.warning("Can't find end of citation inset at line %d!!" %(i))
+            i += 1
+            continue
+        pres = find_token(document.body, "pretextlist", i, j)
+        posts = find_token(document.body, "posttextlist", i, j)
+        if pres == -1 and posts == -1:
+            # nothing to do.
+            i = j + 1
+            continue
+        pretexts = get_quoted_value(document.body, "pretextlist", pres)
+        posttexts = get_quoted_value(document.body, "posttextlist", posts)
+        k = find_token(document.body, "LatexCommand", i, j)
+        if k == -1:
+            document.warning("Can't find LatexCommand for citation inset at line %d!" %(i))
+            i = j + 1
+            continue
+        cmd = get_value(document.body, "LatexCommand", k)
+        if biblatex and cmd in list(ql_citations.keys()):
+            pre = get_quoted_value(document.body, "before", i, j)
+            post = get_quoted_value(document.body, "after", i, j)
+            key = get_quoted_value(document.body, "key", i, j)
+            if not key:
+                document.warning("Citation inset at line %d does not have a key!" %(i))
+                key = "???"
+            keys = key.split(",")
+            prelist = pretexts.split("\t")
+            premap = dict()
+            for pp in prelist:
+                ppp = pp.split(" ", 1)
+                premap[ppp[0]] = ppp[1]
+            postlist = posttexts.split("\t")
+            postmap = dict()
+            for pp in postlist:
+                ppp = pp.split(" ", 1)
+                postmap[ppp[0]] = ppp[1]
+            # Replace known new commands with ERT
+            if "(" in pre or ")" in pre:
+                pre = "{" + pre + "}"
+            if "(" in post or ")" in post:
+                post = "{" + post + "}"
+            res = "\\" + ql_citations[cmd]
+            if pre:
+                res += "(" + pre + ")"
+            if post:
+                res += "(" + post + ")"
+            elif pre:
+                res += "()"
+            for kk in keys:
+                if premap.get(kk, "") != "":
+                    res += "[" + premap[kk] + "]"
+                if postmap.get(kk, "") != "":
+                    res += "[" + postmap[kk] + "]"
+                elif premap.get(kk, "") != "":
+                    res += "[]"
+                res += "{" + kk + "}"
+            document.body[i:j+1] = put_cmd_in_ert([res])
+        else:
+            # just remove the params
+            del document.body[posttexts]
+            del document.body[pretexts]
+            i += 1
+
+
+command_insets = ["bibitem", "citation", "href", "index_print", "nomenclature"]
+def convert_literalparam(document):
+    " Add param literal "
+
+    # These already had some sort of latexify method
+    latexified_insets = ["href", "index_print", "nomenclature"]
+
+    for inset in command_insets:
+        i = 0
+        while True:
+            i = find_token(document.body, '\\begin_inset CommandInset %s' % inset, i)
+            if i == -1:
+                break
+            j = find_end_of_inset(document.body, i)
+            if j == -1:
+                document.warning("Malformed LyX document: Can't find end of %s inset at line %d" % (inset, i))
+                i += 1
+                continue
+            while i < j and document.body[i].strip() != '':
+                i += 1
+            if inset in latexified_insets:
+                document.body.insert(i, "literal \"false\"")
+            else:
+                document.body.insert(i, "literal \"true\"")
+
+
+
+def revert_literalparam(document):
+    " Remove param literal "
+
+    for inset in command_insets:
+        i = 0
+        while True:
+            i = find_token(document.body, '\\begin_inset CommandInset %s' % inset, i)
+            if i == -1:
+                break
+            j = find_end_of_inset(document.body, i)
+            if j == -1:
+                document.warning("Malformed LyX document: Can't find end of %s inset at line %d" % (inset, i))
+                i += 1
+                continue
+            k = find_token(document.body, 'literal', i, j)
+            if k == -1:
+                i += 1
+                continue
+            del document.body[k]
+
+
+
+def revert_multibib(document):
+    " Revert multibib support "
+
+    # 1. Get cite engine
+    engine = "basic"
+    i = find_token(document.header, "\\cite_engine", 0)
+    if i == -1:
+        document.warning("Malformed document! Missing \\cite_engine")
+    else:
+        engine = get_value(document.header, "\\cite_engine", i)
+
+    # 2. Do we use biblatex?
+    biblatex = False
+    if engine in ["biblatex", "biblatex-natbib"]:
+        biblatex = True
+
+    # 3. Store and remove multibib document header
+    multibib = ""
+    i = find_token(document.header, "\\multibib", 0)
+    if i != -1:
+        multibib = get_value(document.header, "\\multibib", i)
+        del document.header[i]
+
+    if not multibib:
+        return
+
+    # 4. The easy part: Biblatex
+    if biblatex:
+        i = find_token(document.header, "\\biblio_options", 0)
+        if i == -1:
+            k = find_token(document.header, "\\use_bibtopic", 0)
+            if k == -1:
+                # this should not happen
+                document.warning("Malformed LyX document! No \\use_bibtopic header found!")
+                return
+            document.header[k-1 : k-1] = ["\\biblio_options " + "refsection=" + multibib]
+        else:
+            biblio_options = get_value(document.header, "\\biblio_options", i)
+            if biblio_options:
+                biblio_options += ","
+            biblio_options += "refsection=" + multibib
+            document.header[i] = "\\biblio_options " + biblio_options
+
+        # Bibtex insets
+        i = 0
+        while (True):
+            i = find_token(document.body, "\\begin_inset CommandInset bibtex", i)
+            if i == -1:
+                break
+            j = find_end_of_inset(document.body, i)
+            if j == -1:
+                document.warning("Can't find end of bibtex inset at line %d!!" %(i))
+                i += 1
+                continue
+            btprint = get_quoted_value(document.body, "btprint", i, j)
+            if btprint != "bibbysection":
+                i += 1
+                continue
+            opts = get_quoted_value(document.body, "biblatexopts", i, j)
+            # change btprint line
+            k = find_token(document.body, "btprint", i, j)
+            if k != -1:
+                document.body[k] = "btprint \"btPrintCited\""
+            # Insert ERT \\bibbysection and wrap bibtex inset to a Note
+            pcmd = "bibbysection"
+            if opts:
+                pcmd += "[" + opts + "]"
+            repl = ["\\begin_inset ERT", "status open", "", "\\begin_layout Plain Layout",\
+                    "", "", "\\backslash", pcmd, "\\end_layout", "", "\\end_inset", "", "",\
+                    "\\end_layout", "", "\\begin_layout Standard", "\\begin_inset Note Note",\
+                    "status open", "", "\\begin_layout Plain Layout" ]
+            repl += document.body[i:j+1]
+            repl += ["", "\\end_layout", "", "\\end_inset", "", ""]
+            document.body[i:j+1] = repl
+            j += 27
+
+            i = j + 1
+        return
+
+    # 5. More tricky: Bibtex/Bibtopic
+    k = find_token(document.header, "\\use_bibtopic", 0)
+    if k == -1:
+        # this should not happen
+        document.warning("Malformed LyX document! No \\use_bibtopic header found!")
+        return
+    document.header[k] = "\\use_bibtopic true"
+
+    # Possible units. This assumes that the LyX name follows the std,
+    # which might not always be the case. But it's as good as we can get.
+    units = {
+        "part" : "Part",
+        "chapter" : "Chapter",
+        "section" : "Section",
+        "subsection" : "Subsection",
+        }
+
+    if multibib not in units.keys():
+        document.warning("Unknown multibib value `%s'!" % nultibib)
+        return
+    unit = units[multibib]
+    btunit = False
+    i = 0
+    while (True):
+        i = find_token(document.body, "\\begin_layout " + unit, i)
+        if i == -1:
+            break
+        if btunit:
+            document.body[i-1 : i-1] = ["\\begin_layout Standard",
+                                "\\begin_inset ERT", "status open", "",
+                                "\\begin_layout Plain Layout", "", "",
+                                "\\backslash",
+                                "end{btUnit}", "\\end_layout",
+                                "\\begin_layout Plain Layout", "",
+                                "\\backslash",
+                                "begin{btUnit}"
+                                "\\end_layout", "", "\\end_inset", "", "",
+                                "\\end_layout", ""]
+            i += 21
+        else:
+            document.body[i-1 : i-1] = ["\\begin_layout Standard",
+                                "\\begin_inset ERT", "status open", "",
+                                "\\begin_layout Plain Layout", "", "",
+                                "\\backslash",
+                                "begin{btUnit}"
+                                "\\end_layout", "", "\\end_inset", "", "",
+                                "\\end_layout", ""]
+            i += 16
+        btunit = True
+        i += 1
+
+    if btunit:
+        i = find_token(document.body, "\\end_body", i)
+        document.body[i-1 : i-1] = ["\\begin_layout Standard",
+                                "\\begin_inset ERT", "status open", "",
+                                "\\begin_layout Plain Layout", "", "",
+                                "\\backslash",
+                                "end{btUnit}"
+                                "\\end_layout", "", "\\end_inset", "", "",
+                                "\\end_layout", ""]
     
 
 ##
@@ -1151,10 +1760,24 @@ convert = [
            [523, []],
            [524, []],
            [525, []],
-           [526, []]
+           [526, []],
+           [527, []],
+           [528, []],
+           [529, []],
+           [530, []],
+           [531, []],
+           [532, [convert_literalparam]],
+           [533, []],
           ]
 
 revert =  [
+           [532, [revert_multibib]],
+           [531, [revert_literalparam]],
+           [530, [revert_qualicites]],
+           [529, [revert_bibpackopts]],
+           [528, [revert_citekeyonly]],
+           [527, [revert_biblatex]],
+           [526, [revert_noprefix]],
            [525, [revert_plural_refs]],
            [524, [revert_labelonly]],
            [523, [revert_crimson, revert_cochinealmath]],