Erlang 基础学习 10 Programming with Files

文件操作函数库的组织方式
• file 模块，打开关闭，读写，列目录，等等基本操作
• filename 模块，一致的处理不同平台上的文件名
• filelib 模块，是file模块的一个扩充，大多是用file模块的函数编写
• io 模块，提供一些函数来处理已打开的文件，比如分析文件内容，将格式化的数据写入文件等

读取文件的不同方式
• 文件内容都是由一些字节组成，这些内容是什么含义完全取决于怎么处理
• 例如文件：
{person, "joe" , "armstrong" ,
        [{occupation, programmer},
         {favoriteLanguage, erlang}]}.
{cat, {name, "zorro" },
      {owner, "joe" }}.
• 1. 一次性读入所有的语句
• file:consult(FileName) 假设文件内包含有合法的erlang语法，成功返回{ok,[Term]}，失败返回{error,Reason}
• 2. 一条一条的读出语句
1> {ok, S} = file:open("data1.dat", read).
{ok,<0.36.0>}
2> io:read(S, '').
{ok,{person,"joe",
     "armstrong",
     [{occupation,programmer},{favoriteLanguage,erlang}]}}
3> io:read(S, '').
{ok,{cat,{name,"zorro"},{owner,"joe"}}}
4> io:read(S, '').
eof
5> file:close(S)
• @spec file:open(File, read) => {ok, IoDevice} | {error, Why}
• @spec io:read(IoDevice, Prompt) => {ok, Term} | {error,Why} | eof
• @spec file:close(IoDevice) => ok | {error, Why}
• 3. 一次读入一行
• 使用io:get_line(IoDevice,Prompt) => Line|eof
• 4. 将整个文件一次性读入，得到一个二进制数据
• file:read_file(File)
• 5. 打开文件，进行随机访问
• file:open(File,[read,binary,raw]).
• file:pread(IoDevice, Start, Len)，从IoDevice中读取Start开始的Len个字节，文件从1开始，如果Len大于文件字节总数，返回全部内容

写文件的方式
• 1. 格式化写入
• @spec io:format(IoDevice, Format, Args) -> ok
• Format 写入的格式，例如
• ~n,~p,~w,~s
• 对于~s(格式化字符串)，可以有更复杂的格式，例如
Format                             Result
io:format("|~10s|~n", ["abc"])       |        abc|
io:format("|~-10s|~n", ["abc"])      |abc       |
io:format("|~10.3.+s|~n",["abc"])   |+++++++abc|
io:format("|~10.10.+s|~n",["abc"]) |abc+++++++|
io:format("|~10.7.+s|~n",["abc"])   |+++abc++++|

• 2. 一次性写入整个文件
• file:write_file(File, IO)
• 3. 随机写入，注意，是覆盖原来的内容
1> {ok, S} = file:open("...", [raw,write,binary])
{ok, ...}
2> file:pwrite(S, 10, <<"new">>)
ok
3> file:close(S)
ok

目录操作
• file:list_dir(Dir).
• file:make_dir(Dir).
• file:delete_dir(Dir).删除非空目录

文件操作
• file:copy(File,Dest).
• file:delete(File).


读取文件信息
• file:read_file_info(File).
-record(file_info,
        {size   :: non_neg_integer(),   % Size of file in bytes.
         type   :: 'device' | 'directory' | 'other' | 'regular' | 'symlink',
         access :: 'read' | 'write' | 'read_write' | 'none',
         atime  :: date_time(), % The local time the file was last read:
                                % {{Year, Mon, Day}, {Hour, Min, Sec}}.
         mtime  :: date_time(), % The local time the file was last written.
         ctime  :: date_time(), % The interpretation of this time field
                                % is dependent on operating system.
                                % On Unix it is the last time the file or
                                % or the inode was changed.  On Windows,
                                % it is the creation time.
         mode   :: integer(),           % File permissions.  On Windows,
                                        % the owner permissions will be
                                        % duplicated for group and user.
         links  :: non_neg_integer(),   % Number of links to the file (1 if the
                                        % filesystem doesn't support links).
         major_device :: integer(),     % Identifies the file system (Unix),
                                        % or the drive number (A: = 0, B: = 1)
                                        % (Windows).
         %% The following are Unix specific.
         %% They are set to zero on other operating systems.
         minor_device,                  % Only valid for devices.
         inode  :: integer(),           % Inode number for file.
         uid    :: integer(),           % User id for owner (integer).
         gid    :: integer()}).         % Group id for owner (integer).

一个文件查找的例子
-module(lib_find).
-export([files/3, files/5]).
-import(lists, [reverse/1]).
-include_lib("kernel/include/file.hrl" ).

%% Dir 要查找的目录
%% Re  正则表达式，例如*.mp3
%% Flag 是否要递归
files(Dir, Re, Flag) ->
    Re1 = regexp:sh_to_awk(Re),
    reverse(files(Dir, Re1, Flag, fun(File, Acc) ->[File|Acc] end, [])).

files(Dir, Reg, Recursive, Fun, Acc) ->
    case file:list_dir(Dir) of
        {ok, Files} -> find_files(Files, Dir, Reg, Recursive, Fun, Acc);
        {error, _} -> Acc
    end.

find_files([File|T], Dir, Reg, Recursive, Fun, Acc0) ->
    FullName = filename:join([Dir,File]),
    case file_type(FullName) of
        regular ->
             case regexp:match(FullName, Reg) of
                  {match, _, _} ->
                      Acc = Fun(FullName, Acc0),
                      find_files(T, Dir, Reg, Recursive, Fun, Acc);
                  _ ->
                      find_files(T, Dir, Reg, Recursive, Fun, Acc0)
             end;
        directory ->
             case Recursive of
                  true ->
                      Acc1 = files(FullName, Reg, Recursive, Fun, Acc0),
                      find_files(T, Dir, Reg, Recursive, Fun, Acc1);
                  false ->
                      find_files(T, Dir, Reg, Recursive, Fun, Acc0)
             end;
        error ->
            find_files(T, Dir, Reg, Recursive, Fun, Acc0)
    end;
find_files([], _, _, _, _, A) ->
    A.
file_type(File) ->
    case file:read_file_info(File) of
        {ok, Facts} ->
             case Facts#file_info.type of
                  regular   -> regular;
                  directory -> directory;
                  _         -> error
             end;
        _ ->
            error
    end.