Hitachi Vantara Pentaho Community Forums
Results 1 to 9 of 9

Thread: Input Text file: pb with format (integer with blank)

  1. #1
    Join Date
    Mar 2007
    Posts
    12

    Default Input Text file: pb with format (integer with blank)

    Hi,

    subject: Spoon uses a bad format to read integer from a CSV file.
    kettle version : Spoon version 2.4.1-M2
    Build version : 1071
    Build date : Sat Mar 17 09:23:14 CET 2007

    I saved a file with "text output file" node and I have a integer column for which some values are written as "1 321", so with a blank (default format from spoon).
    When I try to read this file with "Input text file" node, I have a problem with this "blank" in the integer (see trace below). If I give an explicit format (like ### ### ###) for this comumn, it semms that this format is not taken into account since Spoon always uses a date format (yyy/MM/dd HH:mm:ss.SSS) for an integer !

    Christophe

    2007/03/20 11:06:17 - dummy.0 - Démarrage...
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Unexpected error : be.ibridge.kettle.core.exception.KettleException:
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Error converting line
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) :
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Couldn't parse field [RSA_ID] with value [1 321], format [] ldaf=[yyy/MM/dd HH:mm:ss.SSS]
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : For input string: "1 321"
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : be.ibridge.kettle.core.exception.KettleException:
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Error converting line
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) :
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Couldn't parse field [RSA_ID] with value [1 321], format [] ldaf=[yyy/MM/dd HH:mm:ss.SSS]
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : For input string: "1 321"
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) :
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) :
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at be.ibridge.kettle.trans.step.textfileinput.TextFileInput.convertLineToRow(TextFileInput.java:501)
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at be.ibridge.kettle.trans.step.textfileinput.TextFileInput.processRow(TextFileInput.java:939)
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at be.ibridge.kettle.trans.step.textfileinput.TextFileInput.run(TextFileInput.java:1336)
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Caused by: be.ibridge.kettle.core.exception.KettleException:
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Couldn't parse field [RSA_ID] with value [1 321], format [] ldaf=[yyy/MM/dd HH:mm:ss.SSS]
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : For input string: "1 321"
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) :
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at be.ibridge.kettle.trans.step.textfileinput.TextFileInput.convertLineToRow(TextFileInput.java:464)
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : ... 2 more
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Caused by: java.lang.NumberFormatException: For input string: "1 321"
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at java.lang.NumberFormatException.forInputString(NumberFormatException.java:48)
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at java.lang.Long.parseLong(Long.java:412)
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at java.lang.Long.parseLong(Long.java:461)
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at be.ibridge.kettle.trans.step.textfileinput.TextFileInput.convertValue(TextFileInput.java:624)
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at be.ibridge.kettle.trans.step.textfileinput.TextFileInput.convertLineToRow(TextFileInput.java:422)
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : ... 2 more
    2007/03/20 11:06:17 - Extraction depuis fichier.0 - Fin éxécution étape (Entrées=4, Sorties=0, Lues=0, Ecrites=2, Maj=0, Erreurs=1)

  2. #2
    Join Date
    May 2006
    Posts
    4,882

    Default

    Post your transformation, or adapt the one attached to this mail to match your situation.

    In the attached example 2 integers per row are written to an outputfile. One without format, and one with format "0". a space will precede the first one as this is reserved for the sign (e.g. if the value would be smaller than 0). I've not seen integers with spaces in the middle unless you explicitly use a format.

    For the default format in text input I will check later. I doubt it since I would have hit a some time ago probably.

    Regards,
    Sven

    P.S. I see from your log file that you're not using an explicit format in input as the [] after "format" are empty.
    Attached Files Attached Files
    Last edited by sboden; 03-20-2007 at 07:10 AM.

  3. #3
    Join Date
    Mar 2007
    Posts
    12

    Default

    Sven,

    There are links to the data file and the xml file of the transformation.

    In this small datafile the problem appear for the column "ANNEE_FICH" (year) that should be read as an integer(4) (values : "2 004" or "2 005" : The blanks inside integers have been put by spoon by an "text output node"). The problem persists if I try to use explicit format.
    Why this node try to use a "date format" to read an integer ?
    What does it mean " ldaf=[yyy/MM/dd HH:mm:ss.SSS]" in the log below ?


    2007/03/20 14:34:19 - Extraction depuis fichier.0 - Starting to run...
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - Opening file: /home/christo/rsa/extractions/ftp/K50/rsa_extrait.csv
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Unexpected error : be.ibridge.kettle.core.exception.KettleException:
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Error converting line
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) :
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Couldn't parse field [ANNEE_FICH] with value [2 004], format [] ldaf=[yyy/MM/dd HH:mm:ss.SSS]
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : For input string: "2 004"
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : be.ibridge.kettle.core.exception.KettleException:
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Error converting line
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) :
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Couldn't parse field [ANNEE_FICH] with value [2 004], format [] ldaf=[yyy/MM/dd HH:mm:ss.SSS]
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : For input string: "2 004"
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) :
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) :
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at be.ibridge.kettle.trans.step.textfileinput.TextFileInput.convertLineToRow(TextFileInput.java:501)
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at be.ibridge.kettle.trans.step.textfileinput.TextFileInput.processRow(TextFileInput.java:939)
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at be.ibridge.kettle.trans.step.textfileinput.TextFileInput.run(TextFileInput.java:1336)
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Caused by: be.ibridge.kettle.core.exception.KettleException:
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Couldn't parse field [ANNEE_FICH] with value [2 004], format [] ldaf=[yyy/MM/dd HH:mm:ss.SSS]
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : For input string: "2 004"
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) :
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at be.ibridge.kettle.trans.step.textfileinput.TextFileInput.convertLineToRow(TextFileInput.java:464)
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : ... 2 more
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : Caused by: java.lang.NumberFormatException: For input string: "2 004"
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at java.lang.NumberFormatException.forInputString(NumberFormatException.java:48)
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at java.lang.Long.parseLong(Long.java:412)
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at java.lang.Long.parseLong(Long.java:461)
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at be.ibridge.kettle.trans.step.textfileinput.TextFileInput.convertValue(TextFileInput.java:624)
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : at be.ibridge.kettle.trans.step.textfileinput.TextFileInput.convertLineToRow(TextFileInput.java:422)
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : ... 2 more
    2007/03/20 14:34:19 - Extraction depuis fichier.0 - Fin éxécution étape (Entrées=2, Sorties=0, Lues=0, Ecrites=0, Maj=0, Erreurs=1)
    2007/03/20 14:34:19 - dummy.0 - Démarrage...

  4. #4
    Join Date
    May 2006
    Posts
    4,882

    Default

    To be continued in a couple of hours... I've seen other problems as well with text output step.
    The format is contained in "format []", I have to search what ldaf is (it also misses a "y" or so it seems)

    Regards,
    Sven

  5. #5
    Join Date
    Mar 2007
    Posts
    12

    Default

    I also thought to an encoding problem, but UTF-8 is the correct one.
    Christophe

  6. #6
    Join Date
    May 2006
    Posts
    4,882

    Default

    Naa... I hit the same problem in a different way... Open the text output step where you generate the data and go to the group column in the fields tab for the specific field: you have entered a space there somehow. Remove that space and it will be solved.

    Regards,
    Sven

  7. #7
    Join Date
    Mar 2007
    Posts
    12

    Default Use of format in "Outpout" and "Input Text File" nodes

    Initially my post is about the "input text file" node that is not able to read integer that have "blank" (as separaor for thousand).
    This problem is always open.

    Below I described why I have such "blank" caracter in integer field: It is related to the "output text file".

    Sven: I haven't a space in the group column.

    In fact the problem comes from "format" used by spoon (internally) and concerns "Integer" column.

    In the "text output file" node: In the second tab ("Content") you have a checkbox "Insertion rapide (pas de formatage)". If it is checked then the output is OK and I obtain the following output :

    SECTEUR;ANNEE_FICH;FINESS;RSA_VERSION;TRIMESTRE;RSA_INDEX;VERSION_F;VERSION_GENRSA;VERSION_CLASSIF;LU_GHM_CMD;LU_GHM_TYPE;LU_GHM_NUMERO;LU_GHM_COMPLEXITE;LU_GHM_GROUPAGE;LU_GHM_VERSION_CLASSIF;GHM_CMD;GHM_TYPE;GHM_NUMERO;GHM_COMPLEXITE;GHM_CODE_RETOUR;NB_RUM;AGE_AN;AGE_J;SEXE;MODE_ENTREE;PROVENANCE;MOIS;ANNEE;MODE_SORTIE;DESTINATION;TYPE_SEJOUR;DUREE_SEJ_TOT;CODE_GEO;POIDS;NB_SEANCE;IGS2;NUM_GHS;NB_J;FLG_SEJOUR_INF;NB_DIALYSE;NB_ACTE_24Z05Z;NB_ACTE_24Z06Z;NB_ACTE_24Z07Z;TYPE_PRELEV;NB_SUP_REA;NB_SUP_SOINS;NB_SUP_STF;NB_SUP_SRC;NB_SUP_NN1;NB_SUP_NN2;NB_SUP_NN3;FLG_LIT_PAL;DP;DR;NDAS;NA;RESTE;DUREE_SEJ_REA;RSA_ID;PATIENT_ID;SEJOUR_ID
    PUBLIC;2004;010007987;210;4;4400000000;110;E79;9;23;Z;02;Z;000;9;23;Z;2;Z;000;1;80;;1;8;;12;2004;8;;;29;01110;;0;000;7956;0;0;0;0;0;0;0;;;;;;;;;Z515;C211;1;4;J961 DEQP003001GLHF001001ZBQK002001ZCQK002001;0;1;;


    The strings are trimmed (at right) and no "blank" caracter in integer (2004 is written 2004).

    But if this box is not checked then the output is Bad:

    SECTEUR;ANNEE_FICH;FINESS;RSA_VERSION;TRIMESTRE;RSA_INDEX;VERSION_F;VERSION_GENRSA;VERSION_CLASSIF;LU_GHM_CMD;LU_GHM_TYPE;LU_GHM_NUMERO;LU_GHM_COMPLEXITE;LU_GHM_GROUPAGE;LU_GHM_VERSION_CLASSIF;GHM_CMD;GHM_TYPE;GHM_NUMERO;GHM_COMPLEXITE;GHM_CODE_RETOUR;NB_RUM;AGE_AN;AGE_J;SEXE;MODE_ENTREE;PROVENANCE;MOIS;ANNEE;MODE_SORTIE;DESTINATION;TYPE_SEJOUR;DUREE_SEJ_TOT;CODE_GEO;POIDS;NB_SEANCE;IGS2;NUM_GHS;NB_J;FLG_SEJOUR_INF;NB_DIALYSE;NB_ACTE_24Z05Z;NB_ACTE_24Z06Z;NB_ACTE_24Z07Z;TYPE_PRELEV;NB_SUP_REA;NB_SUP_SOINS;NB_SUP_STF;NB_SUP_SRC;NB_SUP_NN1;NB_SUP_NN2;NB_SUP_NN3;FLG_LIT_PAL;DP;DR;NDAS;NA;RESTE;DUREE_SEJ_REA;RSA_ID;PATIENT_ID;SEJOUR_ID
    PUBLIC;2 004;010007987;210;4;4 400 000 000;110;E79;9;23;Z;02;Z;000;9;23;Z;2;Z;000;1;80;;1;8;;12;2004;8;;;29;01110;;0;000;7956;0;0;0;0;0;0;0;;;;;;;;;Z515 ;C211 ;1;4;J961 DEQP003001GLHF001001ZBQK002001ZCQK002001 ;0;1;;

    The strings are not trimmed (at right) and a "blank" appears in integer (2004 is written 2 004).

    For both cases, the field types are the same: Integer and no exolicit format is given.

    So, why Spoon inserts a "blanck" in integer ?
    What is the default format for integer ?

    But if I give an explicit format like "####" for this integer then all becomes again ok for this integer :

    SECTEUR;ANNEE_FICH;FINESS;RSA_VERSION;TRIMESTRE;RSA_INDEX;VERSION_F;VERSION_GENRSA;VERSION_CLASSIF;LU_GHM_CMD;LU_GHM_TYPE;LU_GHM_NUMERO;LU_GHM_COMPLEXITE;LU_GHM_GROUPAGE;LU_GHM_VERSION_CLASSIF;GHM_CMD;GHM_TYPE;GHM_NUMERO;GHM_COMPLEXITE;GHM_CODE_RETOUR;NB_RUM;AGE_AN;AGE_J;SEXE;MODE_ENTREE;PROVENANCE;MOIS;ANNEE;MODE_SORTIE;DESTINATION;TYPE_SEJOUR;DUREE_SEJ_TOT;CODE_GEO;POIDS;NB_SEANCE;IGS2;NUM_GHS;NB_J;FLG_SEJOUR_INF;NB_DIALYSE;NB_ACTE_24Z05Z;NB_ACTE_24Z06Z;NB_ACTE_24Z07Z;TYPE_PRELEV;NB_SUP_REA;NB_SUP_SOINS;NB_SUP_STF;NB_SUP_SRC;NB_SUP_NN1;NB_SUP_NN2;NB_SUP_NN3;FLG_LIT_PAL;DP;DR;NDAS;NA;RESTE;DUREE_SEJ_REA;RSA_ID;PATIENT_ID;SEJOUR_ID
    PUBLIC;2004;010007987;210;4;4 400 000 000;110;E79;9;23;Z;02;Z;000;9;23;Z;2;Z;000;1;80;;1;8;;12;2004;8;;;29;01110;;0;000;7956;0;0;0;0;0;0;0;;;;;;;;;Z515 ;C211 ;1;4;J961 DEQP003001GLHF001001ZBQK002001ZCQK002001 ;0;1;;

    This example shows that it is very important to control format and maybe to give always explicit format, else file generated by a node "output text file" can be unreadable by the node "input text file" !!!

    Wouldn't be interesting to store format of all columns of a datafile in a separate xml file, and to use this xml file as a "reusable" component (a kind of metadata) ?

    For example it should be very interesting to store the format of columns obtained from a table from a database.
    This approach would have solved my problem since my data comes from a database.

    Talend (an other open source ETL) uses this approach with "Schema type" that can be stored in xml file or in the referential).

    Christophe

  8. #8
    Join Date
    Nov 1999
    Posts
    9,729

    Default

    Kettle uses the defaults locale settings of your OS if you don't override any Encoding, Format, Decimal or Grouping settings.
    I'm fairly sure the locale set on your system puts that space into the integer output format. (unless you put a space in the Group column)

    Wouldn't be interesting to store format of all columns of a datafile in a separate xml file, and to use this xml file as a "reusable" component (a kind of metadata) ?
    Talend (an other open source ETL) uses this approach with "Schema type" that can be stored in xml file or in the referential).
    We could add all kinds of things on top of our own re-useable steps and variables substitution systems. However, adding another XML file to the equation shifts the problem to a packaging issue and creates a useability problem along the way. (I don't want to expose Kettle users to XML)
    IMHO the only way to be really sure of an output result is by explicitely specifying the format mask.
    If you click on the Minimal Width button in the Text File Output step dialog, you will notice that we use the format "0" to represent integers. This will always format in a way that Text File Input can read back.

    The truth of the matter is that there are as many preferences as there are people with regards to text files.

    Finally, you should always use the tool you think is best for the job at hand. If you think that the best tool is Talend (really, you're from France, I completely understand ;-)) then you should use that. However, please do me a favor and don't let me know about it in that case.

    If you have other good ideas about how to improve Pentaho Data Integration, please feel free to add a change request and we'll certainly consider it.

    Cheers,

    Matt

  9. #9
    Join Date
    Mar 2007
    Posts
    12

    Default

    Hi Matt

    Many thanks for your reply.
    Indeed, in my local setting a blank was used as separator for thousand (grouping in spoon).
    So I suppressed this blonak, both for number and money (I use Linux 10.1 and KDE 3.5).
    I've verified that no ablank appears in the "Grouping" column of spoon.

    But nothing is modified in the output file of spoon : I always have a blank in my integer.

    I'm not really annoy that a blank appear (some tools like OpenOffice Calc admit this blank).
    But the problem is that Spoon, is not able to read it's generated file !! (the "input text file" is not able to read the "output text file" !!!) although the "input text file" detects that it has to use a blonk in grouping (since this blank is in data).
    Moreover I d'on't understand the filed ldaf with a data format since my field is declared as an an integer in spoon.

    I've the following error:
    Couldn't parse field [ANNEE_FICH] with value [2 004], format [] ldaf=[yyy/MM/dd HH:mm:ss.SSS]
    2007/03/22 17:48:07 - Extraction depuis fichier.0 - ERROR (version 2.4.1-M2, build 1071 from 2007/03/17 09:23:14) : For input string: "2 004"

    Since "text files are very often used in ETL processes, it is very imortant that such nodes are robust.

    Maybe I should log those points in the bug report application ?

    Christophe

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
Privacy Policy | Legal Notices | Safe Harbor Privacy Policy

Copyright © 2005 - 2019 Hitachi Vantara Corporation. All Rights Reserved.