1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
#[macro_use]
extern crate nom;


use std::fmt;
use nom::*;
use std::path::Path;
use std::fs::File;
use std::io::{Read, Seek, SeekFrom};
use std::collections::HashMap;

fn read_bytes<R: Read+Seek>(input: &mut R, start: u64, length: usize) -> Result<Vec<u8>, String> {
    //println!("Want to read from start={} for length={} bytes to end={}", start, length, start+length as u64);
    let mut res_vec = vec![0; length];

    try!(input.seek(SeekFrom::Start(start)).map_err(|_| "couldn't seek".to_string()));

    try!(input.read_exact(&mut res_vec).map_err(|_| "Couldn't read bytes".to_string()));
    Ok(res_vec)
}

named!(parse_header<(i32, i16, i16)>,
   do_parse!(
       take!(1) >>   // level?
       take!(3) >>   // date last modified
       num_recs: le_i32         >>
       bytes_in_header: le_i16  >>
       bytes_in_rec: le_i16     >>
       take!(2) >>  // res. fill w/ zero
       take!(1) >>  // flag: incomplete transaction
       take!(1) >>  // encryption flag
       take!(12) >> // res. multi-user proc.
       take!(1) >>  // prod. mdx flag
       take!(1) >>  // lang. drv. id
       take!(2) >>  // res.

       ( (num_recs, bytes_in_header, bytes_in_rec) )
   )
);

fn parse_field_name(i: &[u8]) -> String {
    // TODO this accepts UTF8, when it should only accept ASCII
    ::std::str::from_utf8(i).unwrap().trim_right_matches('\x00').to_string()
}

named!(parse_field_descriptor<FieldHeader>,
    do_parse!(
        // FIXME use convert name to String here
        name: take!(11) >>
        field_type: take!(1) >>
        take!(4) >>      // res.
        field_length: be_u8 >>
        decimal_count: be_u8 >>
        take!(2) >>      // work area ID
        take!(1) >>      // ex.
        take!(10)>>      // res.
        take!(1) >>      // prod. mdx flag
        ({
            let field_type = ::std::str::from_utf8(field_type).unwrap().to_string().remove(0);
            let field_type = match field_type {
                'N' => FieldType::Numeric,
                'C' => FieldType::Character,
                _ => { panic!("Unknown char {:?}", field_type) },
            };

            FieldHeader{
                name: parse_field_name(name),
                field_type: field_type,
                field_length: field_length,
                decimal_count: decimal_count,
           }
        })
    )
);

#[derive(Debug)]
pub struct DbfFile<R: Read+Seek> {
    _dbf_file_handle: R,
    _fields: Vec<FieldHeader>,
    _num_recs: u32,
    _bytes_in_rec: u16,
}

pub struct DbfRecordIterator<R: Read+Seek> {
    _dbf_file: DbfFile<R>,
    _next_rec: u32,
}


#[derive(Debug, Clone)]
pub enum FieldType {
    Character,
    Numeric,
    // FIXME more types
}

#[derive(Debug)]
pub enum Field {
    Character(String),
    Numeric(f64),
    Null,
    // FIXME more types
}

impl fmt::Display for Field {
    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
        match self {
            &Field::Character(ref s) => write!(f, "{}", s),
            &Field::Numeric(ref n) => write!(f, "{}", n),
            &Field::Null => write!(f, "(NULL)"),
        }
    }
}


#[derive(Debug, Clone)]
pub struct FieldHeader {
    pub name: String,
    pub field_type: FieldType,
    pub field_length: u8,
    pub decimal_count: u8,
}

pub type Record = HashMap<String, Field>;

impl DbfFile<File> {
    pub fn open_file(filename: &Path) -> Self {
        let dbf_file = File::open(filename).unwrap();
        DbfFile::open(dbf_file)
    }
}

impl<R> DbfFile<R> where R: Read+Seek {
    pub fn open(mut dbf_file: R) -> Self where R: Read+Seek {
        let header_bytes = read_bytes(&mut dbf_file, 0, 32).unwrap();
        let (num_recs, bytes_in_header, bytes_in_rec) = parse_header(&header_bytes).to_result().unwrap();
        // -1 is for the \x0D separator
        // last -1 is maybe because of an off by one error? FIXME
        let num_headers = (bytes_in_header - 1) / 32 - 1;

        let fields: Vec<_> = read_bytes(&mut dbf_file, 32, (num_headers*32) as usize).unwrap().chunks(32).map(|b| parse_field_descriptor(b).to_result().unwrap()).collect();

        DbfFile{ _dbf_file_handle: dbf_file, _fields: fields, _num_recs: num_recs as u32, _bytes_in_rec: bytes_in_rec as u16 }
    }

    pub fn record(&mut self, rec_id: u32) -> Option<Record> {
        if rec_id >= self._num_recs {
            return None;
        }

        let header_length = (32 + 32 * self._fields.len() + 2) as u64;
        let bytes = read_bytes(&mut self._dbf_file_handle, header_length + (rec_id as u64*self._bytes_in_rec as u64), self._bytes_in_rec as usize).or_else(|e| {
            if rec_id == self._num_recs - 1 {
                // If there's an error and it's the last record, then for some reason it works if
                // we take one byte less.
                read_bytes(&mut self._dbf_file_handle, header_length + (rec_id as u64*self._bytes_in_rec as u64), self._bytes_in_rec as usize - 1)
            } else {
                Err(e)
            }
        }).unwrap();
        let mut offset: usize = 0;
        let mut fields = HashMap::with_capacity(self._fields.len());

        for field in self._fields.iter() {
            let this_field_bytes: Vec<_> = bytes.iter().skip(offset).take(field.field_length as usize).map(|x| x.clone()).collect();
            offset = offset + field.field_length as usize;

            let this_field_ascii = String::from_utf8(this_field_bytes).unwrap().trim().to_owned();

            // Is this field a Character
            // FIXME gotta be a better way to do this
            let is_char = match field.field_type { FieldType::Character => true, _ => false };

            // Spec says that a string '*' means NULL, but empty strings are also viewed as null by
            // some software
            let is_null = this_field_ascii.chars().nth(0) == Some('*') || (this_field_ascii.len() == 0 && is_char );

            let value = if is_null {
                Field::Null
            } else {
                match field.field_type {
                    FieldType::Character => Field::Character(this_field_ascii),
                    FieldType::Numeric => Field::Numeric(this_field_ascii.parse().unwrap()),
                }
            };

            fields.insert(field.name.clone(), value);
        }

        Some(fields)
    }

    pub fn records(self) -> DbfRecordIterator<R> {
        DbfRecordIterator{ _dbf_file: self, _next_rec: 0 }
    }

    pub fn num_records(&self) -> u32 {
        return self._num_recs
    }

    pub fn headers(&self) -> &Vec<FieldHeader> {
        &self._fields
    }
}

impl<R> DbfRecordIterator<R> where R: Read+Seek {
    pub fn into_inner(self) -> DbfFile<R> {
        self._dbf_file
    }
}

impl<R> Iterator for DbfRecordIterator<R> where R: Read+Seek {
    type Item = Record;

    fn next(&mut self) -> Option<Record> {
        if self._next_rec >= self._dbf_file._num_recs {
            None
        } else {
            let rec = self._dbf_file.record(self._next_rec);
            self._next_rec = self._next_rec + 1;
            rec
        }
    }

    fn size_hint(&self) -> (usize, Option<usize>) {
        (self._dbf_file._num_recs as usize, Some(self._dbf_file._num_recs as usize))
    }
}