metaproxy  1.13.0
url_recipe.cpp
Go to the documentation of this file.
1 /* This file is part of Metaproxy.
2  Copyright (C) Index Data
3 
4 Metaproxy is free software; you can redistribute it and/or modify it under
5 the terms of the GNU General Public License as published by the Free
6 Software Foundation; either version 2, or (at your option) any later
7 version.
8 
9 Metaproxy is distributed in the hope that it will be useful, but WITHOUT ANY
10 WARRANTY; without even the implied warranty of MERCHANTABILITY or
11 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
12 for more details.
13 
14 You should have received a copy of the GNU General Public License
15 along with this program; if not, write to the Free Software
16 Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
17 */
18 
19 #include "config.hpp"
20 
21 #include <yaz/srw.h>
22 #include <boost/regex.hpp>
23 #include <metaproxy/xmlutil.hpp>
24 
25 #include <string.h>
26 
27 namespace mp_xml = metaproxy_1::xml;
28 
29 std::string mp_xml::url_recipe_handle(xmlDoc *doc, std::string recipe)
30 {
31  std::string result;
32  if (recipe.length() == 0)
33  return result;
34 
35  const xmlNode *ptr1 = xmlDocGetRootElement(doc);
36  while (ptr1 && ptr1->type != XML_ELEMENT_NODE)
37  ptr1 = ptr1->next;
38  if (ptr1)
39  ptr1 = ptr1->children;
40 
41  size_t p0 = 0;
42  while (p0 < recipe.length())
43  {
44  bool uri_encode = false;
45  if (p0 < recipe.length() -1 && recipe[p0] == '%'
46  && recipe[p0 + 1] == '{')
47  {
48  p0++;
49  p0++;
50  uri_encode = true;
51  }
52  else if (p0 < recipe.length() -1 && recipe[p0] == '$'
53  && recipe[p0 + 1] == '{')
54  {
55  p0++;
56  p0++;
57  }
58  else
59  {
60  result += recipe[p0];
61  p0++;
62  continue;
63  }
64  int step = 0; // 0=variable, 1=pattern, 2=replacement, 3=mode
65  std::string variable;
66  std::string pattern;
67  std::string replacement;
68  std::string mode;
69  int c_prev = 0;
70  while (p0 < recipe.length() && step < 5)
71  {
72  char c = recipe[p0];
73  int c_check = c;
74  if (c_prev == '\\')
75  c_check = 0;
76 
77  if (c_check == '}')
78  step = 5;
79  else if (step == 0)
80  {
81  if (c_check == '[')
82  step = 1;
83  else
84  variable += c;
85  }
86  else if (c_check == ']')
87  step = 4;
88  else if (step == 1)
89  {
90  if (c_check == '/')
91  step = 2;
92  else
93  pattern += c;
94  }
95  else if (step == 2)
96  {
97  if (c_check == '/')
98  step = 3;
99  else
100  replacement += c;
101  }
102  else if (step == 3)
103  {
104  mode += c;
105  }
106  c_prev = c;
107  p0++;
108  }
109  if (variable.length())
110  {
111  std::string text;
112  size_t offset = 0;
113  size_t md_pos = variable.find_first_of("md-");
114  if (md_pos == 0)
115  offset = 3;
116  const xmlNode *ptr = ptr1;
117  for (; ptr; ptr = ptr->next)
118  if (ptr->type == XML_ELEMENT_NODE
119  && !strcmp((const char *) ptr->name, "metadata"))
120  {
121  const _xmlAttr *attr = ptr->properties;
122  for (; attr; attr = attr->next)
123  if (!strcmp((const char *) attr->name, "type")
124  && attr->children
125  && !strcmp((const char *) attr->children->content,
126  variable.c_str() + offset))
127  {
128  text = mp_xml::get_text(ptr);
129  break;
130  }
131  }
132  if (pattern.length() > 0)
133  {
134  boost::regex::flag_type b_mode = boost::regex::perl;
135  if (mode.find_first_of('i') != std::string::npos)
136  b_mode |= boost::regex::icase;
137  boost::regex e(pattern, b_mode);
138  boost::match_flag_type match_mode = boost::format_first_only;
139  if (mode.find_first_of('g') != std::string::npos)
140  match_mode = boost::format_all;
141  text = regex_replace(text, e, replacement, match_mode);
142  }
143  if (!uri_encode)
144  result += text;
145  else
146  {
147  char src[2];
148  char dst[4];
149  size_t i;
150  for (i = 0; i < text.length(); i++)
151  {
152  src[0] = text[i];
153  src[1] = '\0';
154  yaz_encode_uri_component(dst, src);
155  result += dst;
156  }
157  }
158  }
159  }
160  return result;
161 }
162 
163 
164 /*
165  * Local variables:
166  * c-basic-offset: 4
167  * c-file-style: "Stroustrup"
168  * indent-tabs-mode: nil
169  * End:
170  * vim: shiftwidth=4 tabstop=8 expandtab
171  */
172