Add parsing of holidays, week, and sunday readings.

Fixes #87

1 files changed, 69 insertions, 29 deletions

diff --git a/generate_html_cs.py b/generate_html_cs.py
index 2251af8..e06d2b5 100755
--- a/generate_html_cs.py
+++ b/generate_html_cs.py
@@ -57,21 +57,44 @@ def csv2dict(filename):
     return out_dict
 
 
+def parse_verses(lines):
+    """
+    Having list of <L> elements, make the text out of them (with
+    <BR> elements where necessary).
+    """
+    out = ""
+    logging.debug("lines = %s", lines)
+    logging.debug("lines = len %s", len(lines))
+    if len(lines) > 0:
+        out = lines[0].text
+        if len(lines) > 1:
+            for line in lines[1:]:
+                logging.debug("lines = %s", lines)
+                logging.debug("another line = %s", line.text)
+                if line.text is not None:
+                    out += "<br>\n" + line.text
+    logging.debug("body = %s", out)
+    return out
+
+
 def parse_body(elem):
     """Parse one verse element
-    Example could be (or NT instead of OT):
+    Example could be (or NT instead of OT, or DAY):
       <OT>
        <S b="Ps" ch="91" v="9"/>
        <L>V Hospodinu je tvé útočiště.</L>
        <SL>Žalm 91,9</SL>
       </OT>
     """
+    if len(list(elem)) == 0:
+        return None
+
     wword = {
         'text': ''
     }
     source_elem = elem.find('S')
-    logging.debug("elem = %s", source_elem)
-    logging.debug("elem = attrib %s", source_elem.attrib)
+    logging.debug("S elem = %s", source_elem)
+    logging.debug("S elem = attrib %s", source_elem.attrib)
 
     # biblical reference (computer readable)
     wword['ref_id'] = {
@@ -85,18 +108,7 @@ def parse_body(elem):
     wword['int_ref_id']['book'] = book_abbrs[wword['ref_id']['book']]
 
     # text of the verse
-    verses = list(elem.getiterator("L"))
-    logging.debug("verses = %s", verses)
-    logging.debug("verses = len %s", len(verses))
-    if len(verses) > 0:
-        wword['text'] = verses[0].text
-        if len(verses) > 1:
-            for line in verses[1:]:
-                logging.debug("verses = %s", verses)
-                logging.debug("another line = %s", line.text)
-                if line.text is not None:
-                    wword['text'] += "<br>\n" + line.text
-    logging.debug("body = %s", wword['text'])
+    wword['text'] = parse_verses(list(elem.getiterator("L")))
 
     # references
     ref = elem.find("SL")
@@ -147,19 +159,45 @@ def parse_losung(elem):
 
     return date_id, out
 
-# TODO: instead of having list, have rather a dictionary with keys being
-# dates; then we can have two-pass parsing, once we would parse <LOSUNG>
-# elements, then <DAY> ones and particular weekly/monthly/anniversary
-# readings to the right place.
-#
-# * What should be the key of the dictionary? (year,month,day) tuple?
-# * First step should be just refactoring of the current code to use
-#   dictionary.
-#
-# For <DAY> it is necessary to distinguish, based on 'type' attribute
-# ('week', 'sunday', 'holiday', 'month') and put generated text to the
-# right property of the day object (and then in the template to the
-# right place)
+
+def parse_day(elem, whole_dict):
+    """
+    Parsse <DAY> element.
+
+    Example:
+       <DAY d="1" m="1" type="holiday" name="Nový Rok  ">
+         <S b="Kol" ch="3" v="17"/>
+         <L>Všechno, cokoli mluvíte nebo děláte,</L>
+         <L>čiňte ve jménu Pána Ježíše a skrze něho děkujte Bohu Otci.</L>
+         <SL>Koloským 3,17</SL>
+       </DAY>
+
+       <DAY d="17" m="2" type="sunday" name="Invocavit"
+            meaning="Vzývati mne bude a vyslyším jej. Žalm 91,15" ord="7">
+         <S b="1J" ch="3" v="8"/>
+         <L>Proto se zjevil Syn Boží, aby zmařil činy ďáblovy.</L>
+         <SL>1.Janova 3,8b</SL>
+       </DAY>
+
+       <DAY d="13" m="1" type="week" name="Alianční modlitební týden  ">
+       </DAY>
+    """
+    date_id = (cur_year, int(elem.attrib["m"]), int(elem.attrib["d"]),)
+    out = whole_dict[date_id]
+
+    if elem.attrib['type'] == "holiday":
+        out['holy_name'] = elem.attrib['name'].strip()
+        out['holy_text'] = parse_body(elem)
+    elif elem.attrib['type'] == "sunday":
+        out['sun_name'] = elem.attrib['name'].strip()
+        if 'meaning' in elem.attrib:
+            out['sun_mean'] = elem.attrib['meaning'].strip()
+        out['sun_ord'] = int(elem.attrib['ord'])
+        out['sun_text'] = parse_body(elem)
+    elif elem.attrib['type'] == "week":
+        out['week_title'] = elem.attrib['name'].strip()
+    else:
+        raise ValueError("Unknown DAY type = %s", elem.attrib['type'])
 
 
 def parse_file(filename):
@@ -172,11 +210,13 @@ def parse_file(filename):
         key, text = parse_losung(los)
         article_dict[key] = text
 
+    for los in tree.getiterator("DAY"):
+        parse_day(los, article_dict)
+
     article_list = []
     article_keys = sorted(article_dict.keys())
     for key in article_keys:
         article_list.append(article_dict[key])
-    # FIXME does Jinja2 somehow sort dictionary if it is on the input?
     return template.render(articles=article_list)