matlab-textscan读带格式的数据文件

>> help textscan
 TEXTSCAN Read formatted data from text file or string.
    C = TEXTSCAN(FID,'FORMAT') reads data from an open text file identified
    by FID into cell array C. Use FOPEN to open the file and obtain FID.
    The FORMAT is a string of conversion specifiers enclosed in single
    quotation marks. The number of specifiers determines the number of
    cells in the cell array C.  For more information, see "Format Options."
   
    C = TEXTSCAN(FID,'FORMAT',N) reads data from the file, using the FORMAT
    N times, where N is a positive integer. To read additional data from
    the file after N cycles, call TEXTSCAN again using the original FID.
 
    C = TEXTSCAN(FID,'FORMAT','PARAM',VALUE) accepts one or more
    comma-separated parameter name/value pairs. For a list of parameters
    and values, see "Parameter Options."
 
    C = TEXTSCAN(FID,'FORMAT',N,'PARAM',VALUE) reads data from the
    file, using the FORMAT N times, and using settings specified by pairs
    of PARAM/VALUE arguments.
 
    C = TEXTSCAN(STR,...) reads data from string STR. You can use the
    FORMAT, N, and PARAM/VALUE arguments described above with this syntax.
    However, for strings, repeated calls to TEXTSCAN restart the scan from
    the beginning each time. (To restart a scan from the last position,
    request a POSITION output.  See also Example 3.)
 
    [C, POSITION] = TEXTSCAN(...) returns the file or string position at
    the end of the scan as the second output argument. For a file, this is
    the value that FTELL(FID) would return after calling TEXTSCAN. For a
    string, POSITION indicates how many characters TEXTSCAN read. 
 
    Notes:
 
    When TEXTSCAN reads a specified file or string, it attempts to match
    the data to the format string. If TEXTSCAN fails to convert a data
    field, it stops reading and returns all fields read before the failure.
 
    Format Options:
 
    The FORMAT string is of the form:  %<WIDTH>.<PREC><SPECIFIER>
        <SPECIFIER> is required; <WIDTH> and <PREC> are optional.
        <WIDTH> is the number of characters or digits to read.
        <PREC> applies only to the family of %f specifiers, and specifies
        the number of digits to read to the right of the decimal point.
 
    Supported values for SPECIFIER:
 
        Numeric Input Type   Specifier   Output Class
        ------------------   ---------   ------------
        Integer, signed        %d          int32
                               %d8         int8
                               %d16        int16
                               %d32        int32
                               %d64        int64
        Integer, unsigned      %u          uint32
                               %u8         uint8
                               %u16        uint16
                               %u32        uint32
                               %u64        uint64
        Floating-point number  %f          double
                               %f32        single
                               %f64        double
                               %n          double
 
        TEXTSCAN converts numeric fields to the specified output type
        according to MATLAB rules regarding overflow, truncation, and the
        use of NaN, Inf, and -Inf.  For example, MATLAB represents an
        integer NaN as zero.
 
        TEXTSCAN imports any complex number as a whole into a complex
        numeric field, converting the real and imaginary parts to the
        specified type (such as %d or %f). Do not include embedded white
        space in a complex number.
 
        Character Strings  Specifier  Details
        -----------------  ---------  -------------------------
        Strings              %s       String
                             %q       String, possibly double-quoted
                             %c       Single character, including delimiter
        Pattern-matching     %[...]   Read only characters in the brackets,
                                      until the first nonmatching
                                      character. To include ] in the set,
                                      specify it first: %[]...].
                             %[^...]  Read only characters not in the
                                      brackets, until the first matching
                                      character. To exclude ], specify it
                                      first: %[^]...].
 
        For each character (%c) specifier, TEXTSCAN returns a char array.
        Other string specifiers return a cell array of strings.
 
    Skipping fields or parts of fields:
 
        Specifier  Action Taken
        ---------  ------------
          %*...    Skip the field. TEXTSCAN does not create an output cell.
          %*N...   Ignore N characters of the field, where N is an integer
                   less than or equal to the number of characters in the
                   field.
 
        Alternatively, include literal text to ignore in the specifier.
        For example, 'Level%u8' reads 'Level1' as 1.
 
        TEXTSCAN does not include leading white-space characters in the
        processing of any data fields. When processing numeric data,
        TEXTSCAN also ignores trailing white space.
 
        If you use the default (white space) field delimiter, TEXTSCAN
        interprets repeated white-space characters as a single delimiter.
        If you specify a nondefault delimiter, TEXTSCAN interprets repeated
        delimiter characters as separate delimiters, and returns an empty
        value to the output cell.
 
    Parameter Options:
 
         Parameter      Value                               Default
         ---------      -----                               -------
         BufSize        Maximum string length in bytes      4095
 
         CollectOutput  If true, TEXTSCAN concatenates      0 (false)
                        consecutive output cells with the
                        same data type into a single array.
 
         CommentStyle   Symbol(s) designating text to       None
                        ignore. Specify a single string
                        (such as '%') to ignore characters
                        following the string on the same
                        line. Specify a cell array of two
                        strings (such as {'/*', '*/'}) to
                        ignore characters between the
                        strings. TEXTSCAN checks for
                        comments only at the start of each
                        field, not within a field.
 
         Delimiter      Field delimiter character(s)        White space
 
         EmptyValue     Value to return for empty numeric   NaN
                        fields in delimited files
 
         EndOfLine      End-of-line character               Determined
                                                            from file:
                                                            \n, \r, or \r\n
 
         ExpChars       Exponent characters                 'eEdD'
 
         Headerlines    Number of lines to skip. Includes   0
                        the remainder of the current line.
 
         MultipleDelimsAsOne                                0 (false)
                        If true, TEXTSCAN treats
                        consecutive delimiters as a single
                        delimiter. Only valid if you
                        specify the 'Delimiter' option.
 
         ReturnOnError  Determines behavior when TEXTSCAN   1 (true)
                        fails to read or convert.  If true,
                        TEXTSCAN terminates without error
                        and returns all fields read.  If
                        false, TEXTSCAN terminates with an
                        error and does not return an output
                        cell array.
 
         TreatAsEmpty String(s) in the data file to       None
                        treat as an empty value. Can be a
                        single string or cell array of
                        strings. Only applies to numeric
                        fields.
 
         Whitespace     White-space characters              ' \b\t'
 
    Examples:
 
    Example 1: Read each column of a text file.
        Suppose the text file 'mydata.dat' contains the following:
            Sally Level1 12.34 45 1.23e10 inf Nan Yes 5.1+3i
            Joe   Level2 23.54 60 9e19 -inf  0.001 No 2.2-.5i
            Bill  Level3 34.90 12 2e5   10  100   No 3.1+.1i
 
        Read the file:
            fid = fopen('mydata.dat');
            C = textscan(fid, '%s%s%f32%d8%u%f%f%s%f');
            fclose(fid);
 
        TEXTSCAN returns a 1-by-9 cell array C with the following cells:
            C{1} = {'Sally','Joe','Bill'}            %class cell
            C{2} = {'Level1'; 'Level2'; 'Level3'}    %class cell
            C{3} = [12.34;23.54;34.9]                %class single
            C{4} = [45;60;12]                        %class int8
            C{5} = [4294967295; 4294967295; 200000]  %class uint32
            C{6} = [Inf;-Inf;10]                     %class double
            C{7} = [NaN;0.001;100]                   %class double
            C{8} = {'Yes','No','No'}                 %class cell
            C{9} = [5.1+3.0i; 2.2-0.5i; 3.1+0.1i]    %class double
 
        The first two elements of C{5} are the maximum values for a 32-bit
        unsigned integer, or intmax('uint32').
 
    Example 2: Read a string, truncating each value to one decimal digit.
        str = '0.41 8.24 3.57 6.24 9.27';
        C = textscan(str, '%3.1f %*1d');
       
        TEXTSCAN returns a 1-by-1 cell array C:
            C{1} = [0.4; 8.2; 3.5; 6.2; 9.2]
 
    Example 3: Resume a text scan of a string.
        lyric = 'Blackbird singing in the dead of night';
        [firstword, pos] = textscan(lyric,'%9c', 1);      %first word
        lastpart = textscan(lyric(pos+1:end), '%s');      %remaining text
 
    For additional examples, type "doc textscan" at the command prompt.

 

=========

实验

>> fid=fopen('e:\readme.txt')

fid =

     4

>> c{:}

ans =

    'Apache'

>> c=textscan(fid,'%10s',2)

c =

    {2x1 cell}

>> c{:}

ans =

    'Lucene'
    'README'

>> fclose(fid)

ans =

     0

>>

 

>> c=textscan(fid,'%7.2f32')

c =

    [3x1 single]

>> c{:}

ans =

  1.0e+004 *

    0.1321
    3.2423
    0.1000

>>

跳过某个字段，使用百分号后跟*

>> fid=fopen('e:\readme.txt')

fid =

     7

>> c=textscan(fid,'%7.2f32 %*n')

c =

    [2x1 single]

>> fclose(fid)

ans =

     0

>> c{:}

ans =

  1.0e+003 *

    1.3210
    0.9999

readme.txt内容是：

1321 32423 999.89

 

 test1.txt内容是：

 

资产负债表   
2004年12月31日   
资 产 期末数 负债及所有者权益 期末数
流动资产：  流动负债： 
　　货币资金 6865234.00  　　短期借款 120000.00
　　应收票据 72120.00      应付票据 85500.00
　　应收账款 38050.00  　　应付账款 80200.00
    减：坏账准备  　　预收账款 
    应收账款净额 38050.00  　　应付工资 
　　其他应收款  　　应付福利费 8430.00
　　预付账款 26600.00  　　应交税金 24420.00
　　存 货 281950.00  　　应付股利 34000.00
　　待摊费用 100.00  　　其他应付款 
　　流动资产合计 7284054.00  　　预提费用 1600.00
长期投资：  　　流动负债合计 354150.00
　长期股权投资  长期负债： 
固定资产：  　　长期借款 804800.00
　固定资产原价 2370000.00  　　应付债券 
　减：累计折旧 41200.00  　　长期负债合计 804800.00
　固定资产净值 2328800.00   
  所有者权益： 
无形资产及其他资产：  　实收资本 8900000.00
　无形资产 500000.00  　资本公积 
　长期待摊费用  　盈余公积 28290.40
　其他长期资产  　未分配利润 25613.60
　无形资产及其他资产合计  500000.00  　所有者权益合计  8953904.00
资产总计 10112854.00  负债及所有者权益总计 10112854.00

 

 >> fclose(fid)

ans =

     0

>> fid=fopen('e:\test1.txt')

fid =

     7

>> c1=textscan(fid,'%s',5)

c1 =

    {5x1 cell}

>> c1{:}

ans =

    '资产负债表'
    '2004年12月31日'
    '资'
    '产'
    '期末数'

>> c1=textscan(fid,'%s',5)

c1 =

    {5x1 cell}

>> c1{:}

ans =

    '负债及所有者权益'
    '期末数'
    '流动资产：'
    '流动负债：'
    '　　货币资金'

>> c1=textscan(fid,'%s',5)

c1 =

    {5x1 cell}

>> c1{:}

ans =

    '6865234.00'
    '　　短期借款'
    '120000.00'
    '　　应收票据'
    '72120.00'

>>

 

>> c1=textscan(fid,'%s %d64 %s %d64 %s %d64',1)

将test1.txt的内容改为：

　　货币资金 6865234.00  　　短期借款 120000.00
　　应收票据 72120.00      应付票据 85500.00
　　应收账款 38050.00  　　应付账款 80200.00
    减：坏账准备  　　预收账款 
    应收账款净额 38050.00  　　应付工资 
　　其他应收款  　　应付福利费 8430.00
　　预付账款 26600.00  　　应交税金 24420.00
　　存 货 281950.00  　　应付股利 34000.00
　　待摊费用 100.00  　　其他应付款 
　　流动资产合计 7284054.00  　　预提费用 1600.00

 

读与括号中的字符不匹配的字符，直到遇上第一个匹配符为止

%[^...]
>> c1=textscan(fid,'%s %[^\n]')

c1 =

    {10x1 cell}    {10x1 cell}

>> c1{:}

ans =

    '　　货币资金'
    '　　应收票据'
    '　　应收账款'
    '减：坏账准备'
    '应收账款净额'
    '　　其他应收款'
    '　　预付账款'
    '　　存'
    '　　待摊费用'
    '　　流动资产合计'

ans =

    '6865234.00  　　短期借款 120000.00 '
    '72120.00      应付票据 85500.00 '
    '38050.00  　　应付账款 80200.00 '
    '　　预收账款 '
    '38050.00  　　应付工资 '
    '　　应付福利费 8430.00 '
    '26600.00  　　应交税金 24420.00 '
    '货 281950.00  　　应付股利 34000.00 '
    '100.00  　　其他应付款 '
    '7284054.00  　　预提费用 1600.00 '

>> c1{1}

ans =

    '　　货币资金'
    '　　应收票据'
    '　　应收账款'
    '减：坏账准备'
    '应收账款净额'
    '　　其他应收款'
    '　　预付账款'
    '　　存'
    '　　待摊费用'
    '　　流动资产合计'

>> c1{2}

ans =

    '6865234.00  　　短期借款 120000.00 '
    '72120.00      应付票据 85500.00 '
    '38050.00  　　应付账款 80200.00 '
    '　　预收账款 '
    '38050.00  　　应付工资 '
    '　　应付福利费 8430.00 '
    '26600.00  　　应交税金 24420.00 '
    '货 281950.00  　　应付股利 34000.00 '
    '100.00  　　其他应付款 '
    '7284054.00  　　预提费用 1600.00 '

 

 将text1.txt改为：

　　货币资金,6865234.00 ,　　短期借款,120000.00
　　应收票据,72120.00 ,    应付票据,85500.00
　　应收账款,38050.00 ,　　应付账款,80200.00
    减：坏账准备,,　　预收账款,
    应收账款净额,38050.00 ,　　应付工资,
　　其他应收款,,　　应付福利费,8430.00
　　预付账款,26600.00 ,　　应交税金,24420.00
　　存 货,281950.00 ,　　应付股利,34000.00
　　待摊费用,100.00 ,　　其他应付款,
　　流动资产合计,7284054.00 ,　　预提费用,1600.00

 

体会一下空字段和分隔符

>> fclose(fid)

ans =

     0

>> fid=fopen('e:\test1.txt')

fid =

     7

>> c1=textscan(fid,'%s %d64 %s %d64','delimiter',',','emptyValue',-Inf)

c1 =

    {10x1 cell}    [10x1 int64]    {10x1 cell}    [10x1 int64]

>> c1{:}

ans =

    '　　货币资金'
    '　　应收票据'
    '　　应收账款'
    '减：坏账准备'
    '应收账款净额'
    '　　其他应收款'
    '　　预付账款'
    '　　存 货'
    '　　待摊费用'
    '　　流动资产合计'

ans =

              6865234
                72120
                38050
 -9223372036854775808
                38050
 -9223372036854775808
                26600
               281950
                  100
              7284054

ans =

    '　　短期借款'
    '应付票据'
    '　　应付账款'
    '　　预收账款'
    '　　应付工资'
    '　　应付福利费'
    '　　应交税金'
    '　　应付股利'
    '　　其他应付款'
    '　　预提费用'

ans =

               120000
                85500
                80200
 -9223372036854775808
 -9223372036854775808
                 8430
                24420
                34000
 -9223372036854775808
                 1600

>>

 

再试一次

 > fid=fopen('e:\test1.txt')

fid =

     7

>> c1=textscan(fid,'%s %d64 %s %d64','delimiter',',','emptyValue',NaN)

c1 =

    {10x1 cell}    [10x1 int64]    {10x1 cell}    [10x1 int64]

>> c1{:}

ans =

    '　　货币资金'
    '　　应收票据'
    '　　应收账款'
    '减：坏账准备'
    '应收账款净额'
    '　　其他应收款'
    '　　预付账款'
    '　　存 货'
    '　　待摊费用'
    '　　流动资产合计'

ans =

              6865234
                72120
                38050
                    0
                38050
                    0
                26600
               281950
                  100
              7284054

ans =

    '　　短期借款'
    '应付票据'
    '　　应付账款'
    '　　预收账款'
    '　　应付工资'
    '　　应付福利费'
    '　　应交税金'
    '　　应付股利'
    '　　其他应付款'
    '　　预提费用'

ans =

               120000
                85500
                80200
                    0
                    0
                 8430
                24420
                34000
                    0
                 1600

>>

 

>> fid=fopen('e:\test1.txt')

fid =

     7

>> c1=textscan(fid,'%s %d64 %s %d64','delimiter',',')

c1 =

    {10x1 cell}    [10x1 int64]    {10x1 cell}    [10x1 int64]

>> c1{:}

ans =

    '　　货币资金'
    '　　应收票据'
    '　　应收账款'
    '减：坏账准备'
    '应收账款净额'
    '　　其他应收款'
    '　　预付账款'
    '　　存 货'
    '　　待摊费用'
    '　　流动资产合计'

ans =

              6865234
                72120
                38050
                    0
                38050
                    0
                26600
               281950
                  100
              7284054

ans =

    '　　短期借款'
    '应付票据'
    '　　应付账款'
    '　　预收账款'
    '　　应付工资'
    '　　应付福利费'
    '　　应交税金'
    '　　应付股利'
    '　　其他应付款'
    '　　预提费用'

ans =

               120000
                85500
                80200
                    0
                    0
                 8430
                24420
                34000
                    0
                 1600

>>