xapian-core/tests/api_replacedoc.cc

   1 /* api_replacedoc.cc: tests of document replacing.
   2  *
   3  * Copyright 2009 Richard Boulton
   4  * Copyright 2015 Olly Betts
   5  *
   6  * This program is free software; you can redistribute it and/or
   7  * modify it under the terms of the GNU General Public License as
   8  * published by the Free Software Foundation; either version 2 of the
   9  * License, or (at your option) any later version.
  10  *
  11  * This program is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14  * GNU General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU General Public License
  17  * along with this program; if not, write to the Free Software
  18  * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301
  19  * USA
  20  */
  21
  22 #include <config.h>
  23
  24 #include "api_replacedoc.h"
  25
  26 #include <string>
  27 #include <map>
  28
  29 using namespace std;
  30
  31 #include <xapian.h>
  32 #include "testsuite.h"
  33 #include "testutils.h"
  34
  35 #include "apitest.h"
  36 #include "dbcheck.h"
  37
  38 // Test that positionlists are updated correctly.
  39 DEFINE_TESTCASE(poslistupdate1, positional && writable) {
  40     Xapian::WritableDatabase db = get_writable_database();
  41
  42     Xapian::Document doc;
  43     doc.add_posting("pos", 2);
  44     doc.add_posting("pos", 3);
  45     db.add_document(doc);
  46     db.flush();
  47
  48     TEST_EQUAL(docterms_to_string(db, 1), "Term(pos, wdf=2, pos=[2, 3])");
  49
  50     doc = db.get_document(1);
  51     doc.add_term("pos2");
  52     db.replace_document(1, doc);
  53     db.flush();
  54     TEST_EQUAL(docterms_to_string(db, 1),
  55                "Term(pos, wdf=2, pos=[2, 3]), "
  56                "Term(pos2, wdf=1)");
  57
  58     doc = db.get_document(1);
  59     doc.add_posting("pos3", 1);
  60     doc.add_posting("pos3", 5);
  61     db.replace_document(1, doc);
  62     db.flush();
  63     TEST_EQUAL(docterms_to_string(db, 1),
  64                "Term(pos, wdf=2, pos=[2, 3]), "
  65                "Term(pos2, wdf=1), "
  66                "Term(pos3, wdf=2, pos=[1, 5])");
  67
  68     doc = db.get_document(1);
  69     doc.remove_term("pos");
  70     db.replace_document(1, doc);
  71     db.flush();
  72     TEST_EQUAL(docterms_to_string(db, 1),
  73                "Term(pos2, wdf=1), "
  74                "Term(pos3, wdf=2, pos=[1, 5])");
  75
  76     // Regression test: the old positionlist fragment used to be left lying
  77     // around here.
  78     Xapian::PositionIterator posit(db.positionlist_begin(1, "pos"));
  79     TEST(posit == db.positionlist_end(1, "pos"));
  80
  81     doc = db.get_document(1);
  82     doc.remove_term("pos3");
  83     db.replace_document(1, doc);
  84     db.flush();
  85     TEST_EQUAL(docterms_to_string(db, 1),
  86                "Term(pos2, wdf=1)");
  87
  88     // Regression test: the old positionlist fragment used to be left lying
  89     // around here.
  90     Xapian::PositionIterator posit2(db.positionlist_begin(1, "pos3"));
  91     TEST(posit2 == db.positionlist_end(1, "pos3"));
  92
  93     doc = db.get_document(1);
  94     doc.add_term("pos");
  95     db.replace_document(1, doc);
  96     db.flush();
  97     TEST_EQUAL(docterms_to_string(db, 1),
  98                "Term(pos, wdf=1), "
  99                "Term(pos2, wdf=1)");
 100
 101     return true;
 102 }
 103
 104 static Xapian::Document
 105 basic_doc() {
 106     Xapian::Document doc;
 107     doc.add_term("z0", 0);
 108     doc.add_term("z1", 1);
 109     return doc;
 110 }
 111
 112 static string
 113 basic_docterms() {
 114     return ", Term(z0, wdf=0), Term(z1, wdf=1)";
 115 }
 116
 117 /** Check that changing the wdf of a term in a document works.
 118  */
 119 DEFINE_TESTCASE(modtermwdf1, writable) {
 120     Xapian::WritableDatabase db(get_writable_database());
 121
 122     string bdt(basic_docterms());
 123
 124     // Add a simple document.
 125     Xapian::Document doc1(basic_doc());
 126     doc1.add_term("takeaway", 1);
 127     db.add_document(doc1);
 128
 129     dbcheck(db, 1, 1);
 130     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=1)" + bdt);
 131
 132     // Modify the wdf of an existing document, checking stats before flush.
 133     Xapian::Document doc2(basic_doc());
 134     doc2.add_term("takeaway", 2);
 135     db.replace_document(1, doc2);
 136     dbcheck(db, 1, 1);
 137     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=2)" + bdt);
 138
 139     // Remove a term, and then put it back again.
 140     Xapian::Document doc0(basic_doc());
 141     db.replace_document(1, doc0);
 142     dbcheck(db, 1, 1);
 143     TEST_EQUAL(docterms_to_string(db, 1), bdt.substr(2));
 144     db.replace_document(1, doc1);
 145     dbcheck(db, 1, 1);
 146     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=1)" + bdt);
 147
 148     // Remove a term, flush, then put it back, remove it, and put it back.
 149     // This is to test the handling of items in the change cache.
 150     db.replace_document(1, doc0);
 151     db.flush();
 152     db.replace_document(1, doc2);
 153     db.replace_document(1, doc0);
 154     db.replace_document(1, doc2);
 155     db.flush();
 156     dbcheck(db, 1, 1);
 157     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=2)" + bdt);
 158
 159     // Remove a term, and then put it back again without checking stats.
 160     db.replace_document(1, doc0);
 161     db.replace_document(1, doc2);
 162     dbcheck(db, 1, 1);
 163     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=2)" + bdt);
 164
 165     // Modify a term, and then put it back again without checking stats.
 166     db.replace_document(1, doc1);
 167     db.replace_document(1, doc2);
 168     dbcheck(db, 1, 1);
 169     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=2)" + bdt);
 170
 171     // Modify the wdf of an existing document, checking stats after flush.
 172     Xapian::Document doc3(basic_doc());
 173     doc3.add_term("takeaway", 3);
 174     db.replace_document(1, doc3);
 175     db.flush();
 176     dbcheck(db, 1, 1);
 177     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=3)" + bdt);
 178
 179     // Change a document, without changing its length.
 180     Xapian::Document doc3_diff(basic_doc());
 181     doc3_diff.add_term("takeaways", 3);
 182     db.replace_document(1, doc3_diff);
 183     db.flush();
 184     dbcheck(db, 1, 1);
 185     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaways, wdf=3)" + bdt);
 186
 187     // Put it back.
 188     db.replace_document(1, doc3);
 189     dbcheck(db, 1, 1);
 190     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=3)" + bdt);
 191
 192     // Modify a document taken from the database.
 193     Xapian::Document doc4(db.get_document(1));
 194     Xapian::Document doc3a(db.get_document(1)); // need this one later
 195     doc3a.termlist_count(); // Pull the document termlist into memory.
 196     doc4.add_term("takeaway", 1);
 197     db.replace_document(1, doc4);
 198     dbcheck(db, 1, 1);
 199     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=4)" + bdt);
 200
 201     // Add a document which was previously added and then modified.
 202     doc1.add_term("takeaway", 1);
 203     db.replace_document(1, doc1);
 204     dbcheck(db, 1, 1);
 205     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=2)" + bdt);
 206
 207     // Add back a document which was taken from the database, but never modified.
 208     db.replace_document(1, doc3a);
 209     dbcheck(db, 1, 1);
 210     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=3)" + bdt);
 211
 212     // Add a position to the document.
 213     Xapian::Document doc5(db.get_document(1));
 214     doc5.add_posting("takeaway", 1, 2);
 215     db.replace_document(1, doc5);
 216     dbcheck(db, 1, 1);
 217     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=5, pos=[1])" + bdt);
 218
 219     // Add a position without changing the wdf.
 220     Xapian::Document doc5b(db.get_document(1));
 221     doc5b.add_posting("takeaway", 2, 0);
 222     db.replace_document(1, doc5b);
 223     dbcheck(db, 1, 1);
 224     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=5, pos=[1, 2])" + bdt);
 225
 226     // Delete a position without changing the wdf.
 227     Xapian::Document doc5c(basic_doc());
 228     doc5c.add_posting("takeaway", 2, 5);
 229     db.replace_document(1, doc5c);
 230     dbcheck(db, 1, 1);
 231     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=5, pos=[2])" + bdt);
 232
 233     // Delete the other position without changing the wdf.
 234     Xapian::Document doc5d(basic_doc());
 235     doc5d.add_term("takeaway", 5);
 236     db.replace_document(1, doc5d);
 237     dbcheck(db, 1, 1);
 238     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=5)" + bdt);
 239
 240     // Reduce the wdf to 0, but don't delete the term.
 241     Xapian::Document doc0freq(basic_doc());
 242     doc0freq.add_term("takeaway", 0);
 243     db.replace_document(1, doc0freq);
 244     dbcheck(db, 1, 1);
 245     TEST_EQUAL(docterms_to_string(db, 1), "Term(takeaway, wdf=0)" + bdt);
 246
 247     // Reduce the wdf to 0, and delete the term.
 248     db.replace_document(1, doc0);
 249     dbcheck(db, 1, 1);
 250     TEST_EQUAL(docterms_to_string(db, 1), bdt.substr(2));
 251
 252     // Delete the document.
 253     db.delete_document(1);
 254     dbcheck(db, 0, 1);
 255     TEST_EXCEPTION(Xapian::DocNotFoundError, docterms_to_string(db, 1));
 256     TEST_EQUAL(postlist_to_string(db, "takeaway"), "");
 257     TEST_EXCEPTION(Xapian::DocNotFoundError, docstats_to_string(db, 1));
 258     TEST_EQUAL(termstats_to_string(db, "takeaway"), "tf=0,cf=0");
 259     TEST_EQUAL(db.get_doccount(), 0);
 260     TEST_EQUAL(db.get_avlength(), 0);
 261     TEST_EQUAL(db.get_lastdocid(), 1);
 262
 263     return true;
 264 }