wip

ferenda/sources/legal/se/offtryck.py

@@ -1083,6 +1083,12 @@ def estimate_density(self, series, resolution, bandwidth):
         return bins, scale
 
     def threshold(self, series, resolution=1000, bandwidth=200):
+        # in the degenerate case that we have a single element series,
+        # there is no way to calculate a threshold between "low" and
+        # "high" values. Just return whatever that element is.
+        assert len(series), "Impossible to calculate a KDE threshold for an empty series"
+        if len(series) == 1:
+            return series[0]
         bins, scale = self.estimate_density(series, resolution, bandwidth)      
 
         # find the valley after the first (significant, not less than

ferenda/sources/legal/se/res/xsl/sfs.xsl

@@ -55,11 +55,14 @@
       </xsl:choose>
       <section id="top" class="col-sm-7">
 	<h1><xsl:value-of select="../xhtml:head/xhtml:title"/></h1>
-	<h2>Version: <xsl:value-of select="$version"/></h2>
+	<!--
+	<xsl:if test="$version">
+	  <h2>Version: <xsl:value-of select="$version"/></h2>
+        </xsl:if>
+        -->
 	<xsl:if test="$label or $alternate">
 	  <p class="lead">(<xsl:value-of select="$label"/><xsl:if test="$label and $alternate">, </xsl:if><xsl:value-of select="$alternate"/>)</p>
 	</xsl:if>
-	<xsl:call-template name="docversions"/>
 	<xsl:call-template name="docmetadata"/>
 	<xsl:if test="../../xhtml:head/xhtml:meta[@rel='rinfoex:upphavdAv']">
 	  <div class="ui-state-error">
@@ -127,7 +130,6 @@
     <xsl:variable name="versions" select="$sfsannotations/rdf:Description[@rdf:about=$documenturi]/dcterms:hasVersion/rdf:Description"/>
     <div class="docversions">
       <xsl:if test="$versions">
-	<h2>Jämför med tidigare versioner</h2>
 	<!--
 	<ul>
 	<xsl:for-each select="$versions">
@@ -137,12 +139,16 @@
 	-->
 	<form action="" method="GET">
 	  <input type="hidden" name="diff" value="true"/>
-	  <select name="from">
+	  <select name="from" id="from" onchange="if (this.options[this.selectedIndex] != 'None') {{ this.form.submit(); }}">
+	    <!--<select name="from" id="from" onchange="this.form.submit()">-->
+	    <option value="None">Jämför med tidigare lydelser</option>
 	    <xsl:for-each select="$versions">
-	      <option value="{substring(dcterms:identifier, 5)}"><xsl:value-of select="dcterms:identifier"/></option>
+	      <option value="{substring(dcterms:identifier, 5)}"><xsl:value-of select="dcterms:identifier"/>
+	      <xsl:if test="rpubl:forarbete"> (<xsl:value-of select="rpubl:forarbete/rdf:Description/dcterms:identifier"/>: <xsl:value-of select="rpubl:forarbete/rdf:Description/dcterms:title"/> <xsl:if test="rpubl:ikrafttradandedatum and rpubl:ikrafttradandedatum != 'None'">, ikraft <xsl:value-of select="rpubl:ikrafttradandedatum"/></xsl:if>)</xsl:if>
+	      <!-- fixme: if the version is the first version (x = y in the url .../x/konsolidering/y), state "(ursprunglig lydelse)" -->
+	      </option>
 	    </xsl:for-each>
 	  </select>
-	  <input type="submit" value="Jämför"/>
 	</form>
       </xsl:if>
     </div>
@@ -156,7 +162,10 @@
       <dt>Utfärdad</dt>
       <dd><xsl:value-of select="//xhtml:meta[@property='rpubl:utfardandedatum']/@content"/></dd>
       <dt>Ändring införd</dt>
-      <dd><xsl:value-of select="//xhtml:meta[@property='dcterms:identifier']/@content"/></dd>
+      <dd><xsl:value-of select="//xhtml:meta[@property='dcterms:identifier']/@content"/>
+	<xsl:call-template name="docversions"/>
+
+      </dd>
       <xsl:if test="//xhtml:meta[@property='rpubl:ikrafttradandedatum']/@content">
 	<dt>Ikraft</dt>
 	<dd><xsl:value-of select="//xhtml:meta[@property='rpubl:ikrafttradandedatum']/@content"/></dd>

ferenda/sources/legal/se/sfs.py

@@ -225,6 +225,8 @@ def download_new(self):
         revisit = []
         if 'revisit' in self.config and self.config.revisit:
             last_revisit = self.config.revisit
+            if not isinstance(last_revisit, list):
+                last_revisit = [last_revisit]
             for wanted_sfs_nr in last_revisit:
                 self.log.info('Revisiting %s' % wanted_sfs_nr)
                 try:
@@ -806,11 +808,6 @@ def extract_metadata_header(self, headertext, basefile):
                                              dummyfile=self.store.parsed_path(basefile))
 
             elif key in ('Departement', 'Departement/ myndighet'):
-                # the split is only needed because of SFS 1942:724,
-                # which has "Försvarsdepartementet,
-                # Socialdepartementet"...
-                if "departementet, " in val:
-                    val = val.split(", ")[0]
                 d["dcterms:creator"] = val
             elif (key == 'Ändring införd' and re_sfs(val)):
                 uppdaterad = re_sfs(val).group(1)
@@ -871,7 +868,12 @@ def sanitize_departement(self, val):
             if val == cleaned:
                 break
             val = cleaned
-        return cleaned
+        # SFS 1942:723: has "Försvarsdepartementet, Socialdepartementet",
+        # SFS 1941:846 has "Miljödepartementet Justitiedepartementet"...
+        # ... just grab the first in these rare cases
+        if re.search("departementet,? [A-Z]", val):
+            val = re.split(",? ", val)[0]
+        return val
 
     def polish_metadata(self, attributes, basefile, infer_nodes=True):
         # attributes will be a nested dict with some values being
@@ -1416,7 +1418,7 @@ def parse_entry_summary(self, doc):
         if ikraft:
             summary += "Ikraftträder: %s\n" % ikraft
         if forarb:
-            display = ", ".join([regpost.meta.value(x, DCTERMS.identifier) for x in forarb])
+            display = ", ".join([regpost.meta.value(x, DCTERMS.identifier) for x in forarb if regpost.meta.value(x, DCTERMS.identifier)])
             summary += "Förarbeten: %s\n" % display
         return summary
 
@@ -2081,7 +2083,7 @@ def updated_sfs_key(row, binding, resource_graph):
                 Facet(RPUBL.arsutgava,
                       use_for_toc=True,
                       label="Ordnade efter utgivningsår",
-                      pagetitle='Författningar utgivna %(selected)s',
+                      pagetitle='Författninagr utgivna %(selected)s',
                       key=sfsnrkey,
                       dimension_label="utgiven",
                       selector_descending=True),

ferenda/sources/legal/se/trips.py

@@ -162,11 +162,26 @@ def _extract_text_inner(self, basefile, attachment=None, version=None, encoding=
         if encoding == "utf-8":
             content = soup.find("div", "search-results-content")
             body = content.find("div", "body-text")
-            if not body or not body.string:
-                raise DocumentRemovedError("%s has no body-text" % basefile,
+            if not body:
+                raise DocumentRemovedError("%s has no body-text div" % basefile,
                                            dummyfile=self.store.parsed_path(basefile))
-            body.string = "----------------------------------------------------------------\n\n" + body.string
-            txt = content.text
+            if body.string:
+                body.string = "----------------------------------------------------------------\n\n" + body.string
+                txt = content.text
+            else:
+                if body.text.strip():
+                    # meaning we have text, just not as a single
+                    # textnode within the <div>. This usually means
+                    # that some < > characters have not been escaped
+                    # in the source material, and are misinterpreted
+                    # by BS4. We warn and salvage what we can.
+                    self.log.warning("Plaintext probably contains unescaped angle brackets, causing an incorrect HTML interpretation")
+                    tmptext = body.text
+                    body.clear()
+                    body.string = "----------------------------------------------------------------\n\n" + tmptext
+                    txt = content.text
+                else:
+                    txt = content.text + "----------------------------------------------------------------\n\n"
         else:
             # some archival pages have a different tag structure
             body = soup.find("pre")

lagen/nu/res/patches/sfs/patches/1942/740.patch

@@ -1,6 +1,6 @@
 --- data/sfs/intermediate/1942/740.txt.bz2
 +++ 
-@@ -9594,11 +9594,14 @@ Korrigerat formattering av 54:8
+@@ -9594,11 +9594,14 @@ [version: 2013:876-] Korrigerat formattering av 54:8
  invändningen mot tingsrättens behörighet grundas på en
  omständighet som högre rätt självmant skall beakta vid ett
  överklagande. Lag (1994:1034).

lagen/nu/res/patches/sfs/patches/1979/1152.patch

@@ -1,6 +1,6 @@
 --- data/sfs/intermediate/1979/1152.txt
 +++ /tmp/tmpvv6a7f_6
-@@ -218,12 +218,14 @@ Justeringar för tabellformattering i 2:2
+@@ -218,12 +218,14 @@ [version: 2009:105-] Justeringar för tabellformattering i 2:2
                  lagring av bränsle och annan byggnad för produktionen utgör
                  kraftverksbyggnad.
 

lagen/nu/res/patches/sfs/patches/1988/609.patch

@@ -1,6 +1,6 @@
 --- data/sfs/intermediate/1988/609.txt.bz2
 +++ 
-@@ -160,7 +160,7 @@ Korrigerat SFS-nummer i hänvisningarna till rättshjälpslagen
+@@ -160,7 +160,7 @@ [version: 2018:535-] Korrigerat SFS-nummer i hänvisningarna till rättshjälpslagen
  Lag (2018:535).
 
  5 § Målsägandebiträdet har rätt till ersättning i enlighet med vad

lagen/nu/res/patches/sfs/patches/1997/192.patch

@@ -1,6 +1,6 @@
 --- data/sfs/intermediate/1997/192.txt
 +++ /tmp/tmpy3vunwtx
-@@ -68,7 +68,6 @@ Formattering av 6 a §
+@@ -68,7 +68,6 @@ [version: 2015:972-] Formattering av 6 a §
  styrelse och revisorer samt stadgar som innebär att
  sammanslutningen är öppen. Lag (2009:289).