exif_read_data
说明
array exif_read_data ( string filename [, string sections [, bool arrays [, bool thumbnail]]] )exif_read_data() reads the EXIF headers from a JPEG or TIFF image file. This way you can read meta data generated by digital cameras.
Exif headers tend to be present in JPEG/TIFF images generated by digital cameras, but unfortunately each digital camera maker has a different idea of how to actually tag their images, so you can't always rely on a specific Exif header being present.
Height and Width are computed the same way getimagesize() does so their values must not be part of any header returned. Also, html is a height/width text string to be used inside normal HTML.
When an Exif header contains a Copyright note, this itself can contain two values. As the solution is inconsistent in the Exif 2.10 standard, the COMPUTED section will return both entries Copyright.Photographer and Copyright.Editor while the IFD0 sections contains the byte array with the NULL character that splits both entries. Or just the first entry if the datatype was wrong (normal behaviour of Exif). The COMPUTED will also contain the entry Copyright which is either the original copyright string, or a comma separated list of the photo and editor copyright.
The tag UserComment has the same problem as the Copyright tag. It can store two values. First the encoding used, and second the value itself. If so the IFD section only contains the encoding or a byte array. The COMPUTED section will store both in the entries UserCommentEncoding and UserComment. The entry UserComment is available in both cases so it should be used in preference to the value in IFD0 section.
注: Windows ME/XP can both wipe the Exif headers when connecting to a camera. More information available at http://www.canon.co.jp/Imaging/NOTICE/011214-e.html.
参数
- filename
The name of the image file being read. This cannot be an URL.
- sections
Is a comma separated list of sections that need to be present in file to produce a result array. If none of the requested sections could be found the return value is FALSE.
FILE FileName, FileSize, FileDateTime, SectionsFound COMPUTED html, Width, Height, IsColor, and more if available. Height and Width are computed the same way getimagesize() does so their values must not be part of any header returned. Also, html is a height/width text string to be used inside normal HTML. ANY_TAG Any information that has a Tag e.g. IFD0, EXIF, ... IFD0 All tagged data of IFD0. In normal imagefiles this contains image size and so forth. THUMBNAIL A file is supposed to contain a thumbnail if it has a second IFD. All tagged information about the embedded thumbnail is stored in this section. COMMENT Comment headers of JPEG images. EXIF The EXIF section is a sub section of IFD0. It contains more detailed information about an image. Most of these entries are digital camera related. - arrays
Specifies whether or not each section becomes an array. The sections COMPUTED, THUMBNAIL, and COMMENT always become arrays as they may contain values whose names conflict with other sections.
- thumbnail
When set to TRUE the thumbnail itself is read. Otherwise, only the tagged data is read.
返回值
It returns an associative array where the array indexes are the header names and the array values are the values associated with those headers. If no data can be returned, exif_read_data() will return FALSE.
更新日志
| 版本 | 说明 |
|---|---|
| 4.3.0 | Can read all embedded IFD data including arrays (returned as such). Also the size of an embedded thumbnail is returned in a THUMBNAIL subarray, and can return thumbnails in TIFF format. Also, there is no longer a maximum length for returned values (not until the memory limit has been reached) |
| 4.3.0 | If PHP has mbstring support, the user comment can automatically change encoding. Also, if the user comment uses Unicode or JIS encoding this encoding will automatically be changed according to the exif ini settings in php.ini |
| 4.3.0 | If the image contains any IFD0 data then COMPUTED contains the entry ByteOrderMotorola which is 0 for little-endian (intel) and 1 for big-endian (motorola) byte order. Also, COMPUTED and UserComment no longer only contain the first copyright entry if the datatype was wrong. |
例