ORA-29270: too many open HTTP requests

Question

Hi,
We have a requirement like, we will get bulk data (2 GB) data file with different languages(Russian, chinese, Japanese, German) etc. and we need to load these data into our databse also we have to translate the same.
We are using Oracle 11g.
We had used Oracle httpuritype - And the below mentioned procedure PROCEDURE TEST.PRC_TRANS_GERMAN will call the function - trans_jef
Input parameter for PRC_TRANS_GERMAN - > To_language and from Language. and this procedure will store the translated word in the seperate table.
Input parameter for the function trans_jef - > Word to be translated , from languae, to_language. And this function will return the translated word.
The issue I am facing is when I am passing the table of around 10000 records, around 4000 rows are not getting translated. So when i try to capture the error in exception- the error is
ORA-29273: HTTP request failed
ORA-06512: at "SYS.UTL_HTTP", line 1819
ORA-29270: too many open HTTP requests
When I try to re execute the missed words, they were translated. But they are not getting translated in one go because of the too many http open request.
So could you please let me know how to resolve this ORA-29270 errror. I know we have to end the request and start again. But not sure how to do. Can some one help with this~!!!!
Translator
CREATE OR REPLACE FUNCTION trans_jef (v_RawText    IN VARCHAR2,
                                      v_SrcLang    IN VARCHAR2,
                                      v_DestLang   IN VARCHAR2)
   RETURN CLOB
IS
    v_EscRawText    VARCHAR2 (3000);
   v_XlateUrl      VARCHAR (3000);
   --v_SrcLang Varchar(10) ;-- := 'de';
   --v_DestLang Varchar(10) ; -- := 'en';
   v_XlateResult   CLOB;
   l_result        CLOB;
BEGIN
   DBMS_OUTPUT.ENABLE (10000000);
    v_EscRawText := UTL_URL.escape (v_RawText, TRUE, 'UTF-8');
   v_XlateUrl :=
         'http://translate.google.com/translate_a/t?client=t&text='
      || v_EscRawText
      || '&hl='
      || v_DestLang
      || '&sl='
      || v_SrcLang
      || '&tl='
      || v_DestLang
      || '&ie=UTF-8&oe=UTF-8&multires=1&oc=1&prev=btn&ssel=0&tsel=0&sc=1';
     v_XlateResult := httpuritype (v_XlateUrl).getclob ();
   -- dbms_output.put_line (v_XlateResult);
   l_result := TRIM (BOTH '"' FROM REGEXP_SUBSTR (v_XlateResult, '"[^"]+"'));
    RETURN l_result;
END;
/* Formatted on 07-09-2012 09:21:05 (QP5 v5.215.12089.38647) */
CREATE OR REPLACE PROCEDURE TEST.PRC_TRANS_GERMAN (TO_LANG1     IN VARCHAR2,
                                                   FROM_LANG1   IN VARCHAR2)
IS
   P_CHINESE         VARCHAR2 (3000);
   FROM_LANG         VARCHAR2 (3000);
   FROM_LANG_CODE    VARCHAR2 (3000);
   TO_LANG_CODE      VARCHAR2 (3000);
   FROM_LANG_CODE1   VARCHAR2 (3000);
   TO_LANG_CODE1     VARCHAR2 (3000);
   cnt               NUMBER := 1;
   counter           NUMBER;
   CURSOR C1
   IS
      SELECT RAW_VAL_TXT FROM T_TRANS_VAL;
   VC1               C1%ROWTYPE;
BEGIN
   OPEN C1;
   LOOP
      FETCH C1 INTO VC1;
      EXIT WHEN C1%NOTFOUND;
      BEGIN
         TO_LANG_CODE :=
            CASE TO_LANG1
               WHEN 'English' THEN 'en'
               WHEN 'Chinese Simplified' THEN 'zh-CN'
               WHEN 'Chinese Tradational' THEN 'zh-TW'
               WHEN 'German' THEN 'de'
               WHEN 'Japanese' THEN 'ja'
            --other languages
            END;
         FROM_LANG_CODE :=
            CASE FROM_LANG1
               WHEN 'English' THEN 'en'
               WHEN 'Chinese Simplified' THEN 'zh-CN'
               WHEN 'Chinese Tradational' THEN 'zh-TW'
               WHEN 'German' THEN 'de'
               WHEN 'Japanese' THEN 'ja'
            other languages
            --                 else 'auto'
            END;
         SELECT trans_jef (VC1.RAW_VAL_TXT, FROM_LANG_CODE, TO_LANG_CODE)
           INTO P_CHINESE
           FROM DUAL;
-- same translator function is used. Language used is simplied chinese.
-- We can use traditional Chinese as well ('zh-TW'), which works perfectly.
         FROM_LANG_CODE1 := --decode(from_lang,'en','English','ar','Arabic', from_lang);
            CASE FROM_LANG_CODE
               WHEN 'en' THEN 'English'
               WHEN 'zh-CN' THEN 'Chinese Simplified'
               WHEN 'zh-TW' THEN 'Chinese Tradational'
               WHEN 'de' THEN 'German'
               WHEN 'ja' THEN 'Japanese'
            --other languages
            END;
         TO_LANG_CODE1 := --decode(from_lang,'en','English','ar','Arabic', from_lang);
            CASE TO_LANG_CODE
               WHEN 'en' THEN 'English'
               WHEN 'zh-CN' THEN 'Chinese Simplified'
               WHEN 'zh-TW' THEN 'Chinese Tradational'
               WHEN 'de' THEN 'German'
               WHEN 'ja' THEN 'Japanese'
            --  other languages
            END;
         INSERT INTO t_trans
              VALUES (FROM_LANG_CODE1,
                      VC1.RAW_VAL_TXT,
                      TO_LANG_CODE1,
                      P_CHINESE);
         COMMIT;
      EXCEPTION
         WHEN OTHERS
         THEN
             INSERT INTO excep_trans (P_WORD_EXP, UPDATE_TIME, COUNTER)
                 VALUES (VC1.RAW_VAL_TXT, SYSDATE, cnt);
            INSERT INTO error_message
                 VALUES (DBMS_UTILITY.FORMAT_ERROR_STACK);
             COMMIT;
      END;
    END LOOP;
   CLOSE C1;
END;
/Edited by: 950814 on 6 Sep, 2012 11:16 PM

Answer 1

Simply add the call to your exception hander. E.g.
EXCEPTION
         WHEN OTHERS
         THEN
               UTL_TCP.Close_All_Connections;
               INSERT INTO excep_trans (P_WORD_EXP, UPDATE_TIME, COUNTER)
                 VALUES (VC1.RAW_VAL_TXT, SYSDATE, cnt);
..etc..
END;Also consider using proper coding standards. There is NO modern day programming language standard that says to write (selective) reserved words in uppercase. Write in lowercase. Use camelcase and pascalcase for variable names and procedures/functions/methods. Restrict the use of uppercase to defining identifiers such as compiler macros or constants.
Do yourself a favour and read, understand and apply the basic naming standards as detailed in .Net Guidelines for Names and Code Conventions for the Java Programming Language.