Parse subjects by delimiter for #2627

[benwbrum]

Implements #2627

[coveralls]

Coverage increased (+0.1%) to 80.777% when pulling c6b3ed3 on 2627-parse-subject-hierarchies into 35e1dd0 on development.

[saracarl]

I don't know if this is a problem or not, but it might be:

<category xml:id="C262">
<catDesc>Activities</catDesc>
<category xml:id="C263">
<catDesc>Agricultural</catDesc>
<category xml:id="S55234">
<catDesc>
<term>hauling oats</term>
<note type="categorization">
Categories:
<ab>
<ptr ana="#category #root" target="#C262">Activities</ptr>
--
<ptr ana="#category #leaf" target="#C263">Agricultural</ptr>
</ab>
</note>

Specifically the -- between ptr tags.

[saracarl]

Exporting a ~200 page TEI document is taking forever. Is this new with this PR or not, I'm not sure.

 CACHE Page Load (0.0ms)  SELECT `pages`.* FROM `pages` WHERE `pages`.`work_id` = 14649 ORDER BY position
  ↳ app/views/export/tei.html.erb:284
  CACHE Note Load (0.0ms)  SELECT `notes`.* FROM `notes` WHERE `notes`.`page_id` IN (504373, 504374, 504375, 504376, 504377, 504378, 504379, 504380, 504381, 504382, 504383, 504384, 504385, 504386, 504387, 504388, 504389, 504390, 504391, 504392, 504393, 504394, 504395, 504396, 504397, 504398, 504399, 504400, 504401, 504402, 504403, 504404, 504405, 504406, 504407, 504408, 504409, 504410, 504411, 504412, 504413, 504414, 504415, 504416, 504417, 504418, 504419, 504420, 504421, 504422, 504423, 504424, 504425, 504426, 504427, 504428, 504429, 504430, 504431, 504432, 504433, 504434, 504435, 504436, 504437, 504438, 504439, 504440, 504441, 504442, 504443, 504444, 504445, 504446, 504447, 504448, 504449, 504450, 504451, 504452, 504453, 504454, 504455, 504456, 504457, 504458, 504459, 504460, 504461, 504462, 504463, 504464, 504465, 504466, 504467, 504468, 504469, 504470, 504471, 504472, 504473, 504474, 504475, 504476, 504477, 504478, 504479, 504480, 504481, 504482, 504483, 504484, 504485, 504486, 504487, 504488, 504489, 504490, 504491, 504492, 504493, 504494, 504495, 504496, 504497, 504498, 504499, 504500, 504501, 504502, 504503, 504504, 504505, 504506, 504507, 504508, 504509, 504510, 504511, 504512, 504513, 504514, 504515, 504516, 504517, 504518, 504519, 504520, 504521, 504522, 504523, 504524, 504525, 504526, 504527, 504528, 504529, 504530, 504531, 504532, 504533, 504534, 504535, 504536, 504537, 504538, 504539, 504540, 504541, 504542, 504543, 504544, 504545, 504546, 504547, 504548, 504549, 504550, 504551, 504552, 504553, 504554, 504555, 504556, 504557, 504558, 504559, 504560, 504561, 504562, 504563, 504564, 504565, 504566) ORDER BY created_at
  ↳ app/views/export/tei.html.erb:284
  CACHE IaLeaf Load (0.0ms)  SELECT `ia_leaves`.* FROM `ia_leaves` WHERE `ia_leaves`.`page_id` IN (504373, 504374, 504375, 504376, 504377, 504378, 504379, 504380, 504381, 504382, 504383, 504384, 504385, 504386, 504387, 504388, 504389, 504390, 504391, 504392, 504393, 504394, 504395, 504396, 504397, 504398, 504399, 504400, 504401, 504402, 504403, 504404, 504405, 504406, 504407, 504408, 504409, 504410, 504411, 504412, 504413, 504414, 504415, 504416, 504417, 504418, 504419, 504420, 504421, 504422, 504423, 504424, 504425, 504426, 504427, 504428, 504429, 504430, 504431, 504432, 504433, 504434, 504435, 504436, 504437, 504438, 504439, 504440, 504441, 504442, 504443, 504444, 504445, 504446, 504447, 504448, 504449, 504450, 504451, 504452, 504453, 504454, 504455, 504456, 504457, 504458, 504459, 504460, 504461, 504462, 504463, 504464, 504465, 504466, 504467, 504468, 504469, 504470, 504471, 504472, 504473, 504474, 504475, 504476, 504477, 504478, 504479, 504480, 504481, 504482, 504483, 504484, 504485, 504486, 504487, 504488, 504489, 504490, 504491, 504492, 504493, 504494, 504495, 504496, 504497, 504498, 504499, 504500, 504501, 504502, 504503, 504504, 504505, 504506, 504507, 504508, 504509, 504510, 504511, 504512, 504513, 504514, 504515, 504516, 504517, 504518, 504519, 504520, 504521, 504522, 504523, 504524, 504525, 504526, 504527, 504528, 504529, 504530, 504531, 504532, 504533, 504534, 504535, 504536, 504537, 504538, 504539, 504540, 504541, 504542, 504543, 504544, 504545, 504546, 504547, 504548, 504549, 504550, 504551, 504552, 504553, 504554, 504555, 504556, 504557, 504558, 504559, 504560, 504561, 504562, 504563, 504564, 504565, 504566)
  ↳ app/views/export/tei.html.erb:284
  CACHE ScCanvas Load (0.0ms)  SELECT `sc_canvases`.* FROM `sc_canvases` WHERE `sc_canvases`.`page_id` IN (504373, 504374, 504375, 504376, 504377, 504378, 504379, 504380, 504381, 504382, 504383, 504384, 504385, 504386, 504387, 504388, 504389, 504390, 504391, 504392, 504393, 504394, 504395, 504396, 504397, 504398, 504399, 504400, 504401, 504402, 504403, 504404, 504405, 504406, 504407, 504408, 504409, 504410, 504411, 504412, 504413, 504414, 504415, 504416, 504417, 504418, 504419, 504420, 504421, 504422, 504423, 504424, 504425, 504426, 504427, 504428, 504429, 504430, 504431, 504432, 504433, 504434, 504435, 504436, 504437, 504438, 504439, 504440, 504441, 504442, 504443, 504444, 504445, 504446, 504447, 504448, 504449, 504450, 504451, 504452, 504453, 504454, 504455, 504456, 504457, 504458, 504459, 504460, 504461, 504462, 504463, 504464, 504465, 504466, 504467, 504468, 504469, 504470, 504471, 504472, 504473, 504474, 504475, 504476, 504477, 504478, 504479, 504480, 504481, 504482, 504483, 504484, 504485, 504486, 504487, 504488, 504489, 504490, 504491, 504492, 504493, 504494, 504495, 504496, 504497, 504498, 504499, 504500, 504501, 504502, 504503, 504504, 504505, 504506, 504507, 504508, 504509, 504510, 504511, 504512, 504513, 504514, 504515, 504516, 504517, 504518, 504519, 504520, 504521, 504522, 504523, 504524, 504525, 504526, 504527, 504528, 504529, 504530, 504531, 504532, 504533, 504534, 504535, 504536, 504537, 504538, 504539, 504540, 504541, 504542, 504543, 504544, 504545, 504546, 504547, 504548, 504549, 504550, 504551, 504552, 504553, 504554, 504555, 504556, 504557, 504558, 504559, 504560, 504561, 504562, 504563, 504564, 504565, 504566)
  ↳ app/views/export/tei.html.erb:284

[benwbrum]

Running a 200pg export from this branch takes 3.2 minutes
Running the same export from development takes 2.9 minutes

[benwbrum]

The -- characters between pointer tags are intentional, and not introduced by this PR.