Add support for half-open search terms like 5-.

However, `-5` is a NEGATION search, not a half-open term. Use `1-5`.
kohler · Jan 27, 2024 · 7d6d279 · 7d6d279
1 parent d99a795
commit 7d6d279
Show file tree

Hide file tree

Showing 3 changed files with 26 additions and 8 deletions.
diff --git a/src/papersearch.php b/src/papersearch.php
@@ -727,7 +727,7 @@ private function _kwdef_parse($kwdef, $sword, $kwexplicit) {
     static private function _search_word_is_paperid($str) {
         $ch = substr($str, 0, 1);
         return ($ch === "#" || ctype_digit($ch))
-            && preg_match('/\A(?:#?\d+(?:(?:-|–|—)#?\d+)?(?:\s*,\s*|\z))+\z/s', $str);
+            && preg_match('/\A(?:#?\d++(?:(?:-|–|—)(?:|#?\d++))?(?:\s*,\s*|\z))+\z/s', $str);
     }
 
     /** @param string $str
@@ -763,7 +763,7 @@ private function _search_word($kw, $sword, $scope) {
             }
         }
 
-        // Paper ID search term (`1-2`, `#1-#2`, etc.)
+        // Paper ID search term (`1-2`, `#1-#2`, `1-`, etc.)
         if (!$sword->quoted
             && !$scope->defkw
             && self::_search_word_is_paperid($sword->word)) {

diff --git a/src/searchterm.php b/src/searchterm.php
@@ -1467,7 +1467,11 @@ private function add_drange($p0, $p1, $rev, $explicit) {
                     $n = $this->n + ($rev ? $p1x - $p0 - 1 : 0);
                     array_splice($this->r, $i, 0, [[$p0, $p1x, $n, $rev, $explicit]]);
                 }
-                $this->n += $p1x - $p0;
+                // ensure `$this->n <= PHP_INT_MAX`
+                // (it naturally will be, UNLESS someone calls add_range
+                // with negative PIDs)
+                $delta = min($p1x - $p0, PHP_INT_MAX - $this->n);
+                $this->n += $delta;
             }
             $p0 = max($p0, $p1x);
         }
@@ -1524,7 +1528,7 @@ function sql_predicate($field) {
             return "false";
         } else if ($this->n <= 8 * count($this->r)
                    && ($pids = $this->paper_ids()) !== null) {
-            return "$field in (" . join(",", $pids) . ")";
+            return "{$field} in (" . join(",", $pids) . ")";
         } else {
             $s = [];
             foreach ($this->r as $r) {
@@ -1568,10 +1572,18 @@ static function parse_pidcode($word, SearchWord $sword, PaperSearch $srch) {
      * @return PaperID_SearchTerm */
     static function parse_normal($word) {
         $st = new PaperID_SearchTerm;
-        while (preg_match('/\A#?(\d+)(?:(?:-|–|—)#?(\d+))?\s*,?\s*(.*)\z/s', $word, $m)) {
-            $m[2] = (isset($m[2]) && $m[2] ? $m[2] : $m[1]);
-            $st->add_range(intval($m[1]), intval($m[2]));
-            $word = $m[3];
+        $pos = 0;
+        while (preg_match('/\G#?(\d++)((?:-|–|—)#?(\d++)|(?:-|–|—)|)\s*,?\s*/s', $word, $m, 0, $pos)) {
+            $p1 = intval($m[1]);
+            if ($m[2] === "") {
+                $p2 = $p1;
+            } else if (!isset($m[3]) || $m[3] === "") {
+                $p2 = PHP_INT_MAX;
+            } else {
+                $p2 = intval($m[3]);
+            }
+            $st->add_range($p1, $p2);
+            $pos += strlen($m[0]);
         }
         return $st;
     }

diff --git a/test/t_search.php b/test/t_search.php
@@ -75,6 +75,12 @@ function test_xor() {
         xassert_search($this->conf->root_user(), "1-10 XOR 4-5", "1 2 3 6 7 8 9 10");
     }
 
+    function test_halfopen_interval() {
+        xassert_search($this->conf->root_user(), "5-100000 XOR 10-100000", "5 6 7 8 9");
+        xassert_search($this->conf->root_user(), "5- XOR 10-100000", "5 6 7 8 9");
+        xassert_search($this->conf->root_user(), "8-,7-,6-,5- XOR 10-100000", "5 6 7 8 9");
+    }
+
     function test_review_term_to_round_mask() {
         $rl = $this->conf->round_list();
         xassert_eqq($rl[0], "");