-rw-r--r-- | lib/data.cc | 3 | ||||
-rw-r--r-- | lib/discovery.cc | 17 | ||||
-rw-r--r-- | lib/util.cc | 41 |
3 files changed, 49 insertions, 12 deletions
diff --git a/lib/data.cc b/lib/data.cc index c040430..f71788f 100644 --- a/lib/data.cc +++ b/lib/data.cc | |||
@@ -9,2 +9,5 @@ namespace opkele { | |||
9 | 9 | ||
10 | const char *_iname_leaders = "=@+$!("; | ||
11 | const char *_whitespace_chars = " \t\r\n"; | ||
12 | |||
10 | } | 13 | } |
diff --git a/lib/discovery.cc b/lib/discovery.cc index b7f2db6..5913ad4 100644 --- a/lib/discovery.cc +++ b/lib/discovery.cc | |||
@@ -8,2 +8,3 @@ | |||
8 | #include <opkele/tidy.h> | 8 | #include <opkele/tidy.h> |
9 | #include <opkele/data.h> | ||
9 | #include <opkele/debug.h> | 10 | #include <opkele/debug.h> |
@@ -23,4 +24,2 @@ namespace opkele { | |||
23 | 24 | ||
24 | static const char *whitespace = " \t\r\n"; | ||
25 | static const char *i_leaders = "=@+$!("; | ||
26 | static const size_t max_html = 16384; | 25 | static const size_t max_html = 16384; |
@@ -130,6 +129,6 @@ namespace opkele { | |||
130 | idiscovery_t idis; | 129 | idiscovery_t idis; |
131 | string::size_type fsc = identity.find_first_not_of(whitespace); | 130 | string::size_type fsc = identity.find_first_not_of(data::_whitespace_chars); |
132 | if(fsc==string::npos) | 131 | if(fsc==string::npos) |
133 | throw bad_input(OPKELE_CP_ "whitespace-only identity"); | 132 | throw bad_input(OPKELE_CP_ "whitespace-only identity"); |
134 | string::size_type lsc = identity.find_last_not_of(whitespace); | 133 | string::size_type lsc = identity.find_last_not_of(data::_whitespace_chars); |
135 | assert(lsc!=string::npos); | 134 | assert(lsc!=string::npos); |
@@ -141,3 +140,3 @@ namespace opkele { | |||
141 | idis.clear(); | 140 | idis.clear(); |
142 | if(strchr(i_leaders,id[0])) { | 141 | if(strchr(data::_iname_leaders,id[0])) { |
143 | /* TODO: further normalize xri identity? Like folding case | 142 | /* TODO: further normalize xri identity? Like folding case |
@@ -497,3 +496,3 @@ namespace opkele { | |||
497 | href.assign(ns); | 496 | href.assign(ns); |
498 | string::size_type lns=href.find_last_not_of(whitespace); | 497 | string::size_type lns=href.find_last_not_of(data::_whitespace_chars); |
499 | href.erase(lns+1); | 498 | href.erase(lns+1); |
@@ -501,5 +500,5 @@ namespace opkele { | |||
501 | } | 500 | } |
502 | for(string::size_type ns=rels.find_first_not_of(whitespace); | 501 | for(string::size_type ns=rels.find_first_not_of(data::_whitespace_chars); |
503 | ns!=string::npos; ns=rels.find_first_not_of(whitespace,ns)) { | 502 | ns!=string::npos; ns=rels.find_first_not_of(data::_whitespace_chars,ns)) { |
504 | string::size_type s = rels.find_first_of(whitespace,ns); | 503 | string::size_type s = rels.find_first_of(data::_whitespace_chars,ns); |
505 | string rel; | 504 | string rel; |
diff --git a/lib/util.cc b/lib/util.cc index bb8a2e8..29e6738 100644 --- a/lib/util.cc +++ b/lib/util.cc | |||
@@ -15,2 +15,3 @@ | |||
15 | #include <opkele/exception.h> | 15 | #include <opkele/exception.h> |
16 | #include <opkele/data.h> | ||
16 | #include <opkele/debug.h> | 17 | #include <opkele/debug.h> |
@@ -213,5 +214,4 @@ namespace opkele { | |||
213 | string rfc_3986_normalize_uri(const string& uri) { | 214 | string rfc_3986_normalize_uri(const string& uri) { |
214 | static const char *whitespace = " \t\r\n"; | ||
215 | string rv; | 215 | string rv; |
216 | string::size_type ns = uri.find_first_not_of(whitespace); | 216 | string::size_type ns = uri.find_first_not_of(data::_whitespace_chars); |
217 | if(ns==string::npos) | 217 | if(ns==string::npos) |
@@ -225,3 +225,3 @@ namespace opkele { | |||
225 | bool s; | 225 | bool s; |
226 | string::size_type ul = uri.find_last_not_of(whitespace)+1; | 226 | string::size_type ul = uri.find_last_not_of(data::_whitespace_chars)+1; |
227 | if(ul <= (colon+3)) | 227 | if(ul <= (colon+3)) |
@@ -432,2 +432,37 @@ namespace opkele { | |||
432 | 432 | ||
433 | string normalize_identifier(const string& usi,bool strip_fragment) { | ||
434 | if(usi.empty()) | ||
435 | return usi; | ||
436 | string rv; | ||
437 | string::size_type fsc = usi.find_first_not_of(data::_whitespace_chars); | ||
438 | if(fsc==string::npos) | ||
439 | return rv; | ||
440 | string::size_type lsc = usi.find_last_not_of(data::_whitespace_chars); | ||
441 | assert(lsc!=string::npos); | ||
442 | if(!strncasecmp(usi.c_str()+fsc,"xri://",sizeof("xri://")-1)) | ||
443 | fsc += sizeof("xri://")-1; | ||
444 | if( (fsc+1) >= lsc ) | ||
445 | return rv; | ||
446 | rv.assign(usi,fsc,lsc-fsc+1); | ||
447 | if(strchr(data::_iname_leaders,rv[0])) { | ||
448 | /* TODO: further normalize xri identity, fold case or | ||
449 | * whatever... */ | ||
450 | }else{ | ||
451 | if(rv.find("://")==string::npos) | ||
452 | rv.insert(0,"http://"); | ||
453 | if(strip_fragment) { | ||
454 | string::size_type fp = rv.find('#'); | ||
455 | if(fp!=string::npos) { | ||
456 | string::size_type qp = rv.find('?'); | ||
457 | if(qp==string::npos || qp<fp) | ||
458 | rv.erase(fp); | ||
459 | else if(qp>fp) | ||
460 | rv.erase(fp,qp-fp); | ||
461 | } | ||
462 | } | ||
463 | rv = rfc_3986_normalize_uri(rv); | ||
464 | } | ||
465 | return rv; | ||
466 | } | ||
467 | |||
433 | } | 468 | } |