![[BACK]](/cvsweb/icons/back.gif){kind=icon}
Return to writeup.tx CVS log ![[TXT]](/cvsweb/icons/text.gif){kind=icon}
![[DIR]](/cvsweb/icons/dir.gif){kind=icon}
Up to [CSRG BSD Unix] / 3BSD / new / libI77uc|
Return to writeup.tx CVS log | Up to [CSRG BSD Unix] / 3BSD / new / libI77uc |
1.1 root 1: .ND
2: .TL
3: Update to the f77 I/O Library
4: January 1980
5: .AU
6: David L. Wasley
7: .AI
8: University of California
9: Berkeley, Calif. 94720
10: .PP
11: The fortran-77 I/O library, libI77.a, has been extensively updated.
12: All known bugs have been fixed, and
13: I/O error reporting has been improved.
14: Several non-standard extensions to FORTRAN I/O have been added.
15: .PP
16: Some general concepts regarding f77 I/O deserve clarification. There are three
17: forms of I/O: formatted, unformatted, and list-directed. The last is
18: related to formatted but does not obey all the rules for formatted I/O.
19: There are two modes of access to external and internal files: direct
20: and sequential. The definition of a logical record depends upon the
21: combination of I/O form and mode specified by the fortran I/O statement.
22: .PP
23: A logical record in direct access external files is a string of bytes
24: of a length specified when the file is opened.
25: Read and write statements must not specify logical records longer than
26: the original record size definition. Shorter logical records are allowed.
27: Unformatted direct writes leave the unfilled part of the record undefined.
28: Formatted direct writes cause the unfilled record to be padded with blanks.
29: .PP
30: Logical records in sequentially accessed external files may be of arbitrary
31: and variable length.
32: Logical record length for unformatted sequential files is determined by
33: the size of items in the iolist.
34: For formatted write statements, logical record length is determined by
35: the format statement interacting with the iolist at execution time.
36: Formatted sequential access causes one or more logical records
37: ending with 'newline' characters to be read or written.
38: .PP
39: Logical record length for list-directed I/O is relatively meaningless.
40: On output, the record length is dependent on the magnitude of the
41: data items.
42: On input, the record length is determined by the data types and the file
43: contents.
44: .PP
45: The logical record length for "internal" files is the length of the
46: character variable or array element. Thus a simple character variable
47: is a single logical record. A character variable array is similar to
48: a fixed length direct access file, and obeys the same rules.
49: Unformatted I/O is not allowed on "internal" files.
50: .PP
51: Note that each execution of a fortran unformatted I/O statement causes a single
52: logical record to be read or written. Each execution of a fortran formatted
53: I/O statement causes one or more logical records to be read or written.
54: .PP
55: Any error detected during I/O processing will cause the program to abort
56: unless alternate action has been provided for specifically in the program.
57: Any I/O statement may include an err= clause (and iostat= clause)
58: to specify an
59: alternate branch to be taken on errors (and return the specific error code).
60: Read or write statements may include end= to branch on end-of-file.
61: File position and the value of I/O list items is undefined following an error.
62:
63: I. Implementation details.
64: .PP
65: The maximum number of logical units that a program may have open at one
66: time has been set to correspond with the UNIX system limit, currently 20.
67: However, the I/O library uses UNIX file access for internal purposes.
68: Therefore fatal errors are possible if the maximum number of files are open.
69: Specifically, 'close' or 'endfile' on an old file,
70: and "'inquire' by file" may fail.
71: .PP
72: Vertical format control is implemented. The logical unit must be opened
73: for sequential access and "form = 'print'" (see below).
74: Control codes '0' and '1' are replaced in the output file
75: with '\\n' and '\\f' respectively.
76: The control character '+' isn't implemented and, like
77: any other character in the first position of a record
78: written to a "print" file, is dropped.
79: No vertical format control is recognized for direct formatted output
80: or list directed output.
81: .PP
82: Default logical units 0, 5, and 6 can be re-defined with an 'open' statement.
83: To preserve error reporting, it is an error to close logical unit 0.
84: If you want to open the default filename for any preconnected logical unit,
85: remember to 'close' the unit first.
86: Redefining the standard units may impair normal console I/O.
87: An alternative is to
88: use shell re-direction to externally re-define the above units.
89: To re-define default blank control or format of the standard input or output
90: files, use the 'open' statement specifying the unit number and no
91: filename (see below).
92: .PP
93: An 'open' statement need not specify a filename. If it refers to a logical
94: unit that is already open, the "blank= " and "form= " specifiers may be
95: redefined without affecting the current file position.
96: Otherwise, if "status='scratch'" is specified, a temporary file with a
97: name of the form 'tmp.FXXXXXX' will be opened,
98: and, by default, will be deleted when closed or during
99: termination of program execution.
100: Any other "status= " specifier without an associated filename results in
101: opening a file named 'fort.N' where N is the specified logical unit number.
102: It is an error to try to open an existing file with "status='new'".
103: It is an error to try to open a nonexistent file with "status='old'".
104: By default "status='unknown'" will be assumed, and a file will be created
105: if necessary.
106: Existing files are never truncated on opening but are positioned
107: at the end-of-file.
108: .PP
109: Sequentially accessed external files are truncated to the current file
110: position on 'close', 'backspace', or 'rewind' only if the last
111: access to the file was a write.
112: .PP
113: Upper as well as lower case characters are recognized in format statements
114: and all alphabetic arguments to the I/O library routines.
115: This has always been true for statements that are
116: part of the source code, but not for format statements
117: or character arguments from a file.
118: .PP
119: If the external representation of a datum
120: is too large for the field width specified, the specified
121: field is filled with asterisks (*).
122: On 'Ew.dEe' output, the e field will be filled with asterisks if the
123: exponent representation is too large.
124: (This will only happen if e==0)
125: .PP
126: List-directed output of complex values now includes an appropriate comma.
127: List-directed output now distinguishes between real*4 and real*8 values
128: and formats them differently.
129: Output of a character string that includes '\\n' now works correctly.
130: .PP
131: If I/O errors are not trapped by the user's program an appropriate
132: error message will be written to 'stderr' before aborting.
133: An error number will be printed in [ ] along with a brief error message
134: showing the logical unit and I/O state.
135: Error numbers < 100 refer to UNIX errors, and are described in the
136: introduction to chapter 2 of the UNIX Programmer's Manual.
137: Error numbers >= 100 come from the I/O library, and are described
138: further in the appendix to this writeup.
139: For internal I/O, part of the string will be printed with '|' at the
140: current position in the string.
141: For external I/O, part of the current record will be displayed if
142: the error was caused during reading from a file that can backspace.
143: .PP
144: Direct access list-directed I/O is not allowed.
145: Unformatted internal I/O is not allowed.
146: Both the above will be caught by the compiler.
147: All other flavors of I/O are allowed, although some are not part of the ANSI
148: standard.
149: .PP
150: The standard units, 0, 5, and 6, are now named internally 'stderr', 'stdin',
151: and 'stdout' respectively.
152: These are not actual filenames and can not be used for opening these units.
153: \'inquire' will not return these names and will indicate
154: that the above units are not named unless they have been opened to real files.
155: The names are meant to make error reporting more meaningful.
156: .PP
157: On output, a real value that is truly zero will display as '0.' to
158: distinguish it from a very small non-zero value.
159: This occurs in 'F', 'E', 'D', and 'G' format conversions.
160: .PP
161: Non-destructive tabbing is implemented for both internal and external
162: formatted I/O.
163: Tabbing left or right on output
164: does not affect previously written portions of a record.
165: Tabbing right on output
166: causes unwritten portions of a record to be filled with blanks.
167: Tabbing left or right off the end of a logical record is an error.
168: The format specifier 'T' must be followed by a positive non-zero number.
169: If it is not, it will have a different meaning (See below).
170: Note that spacing with 'X' always writes blanks in the output record.
171:
172: II. Non-"ANSI Standard" Extensions
173: .PP
174: B is an acceptable edit control specifier. It causes return to the
175: default mode of blank interpretation (NULL) and is identical to BN.
176: This is consistent with S which returns to default sign control.
177: .PP
178: P by itself is equivalent to 0P. It resets the scale factor to the
179: default value, 0.
180: .PP
181: The form of the 'Ew.dEe' format specifier has been extended to 'D' also.
182: The form 'Ew.d.e' is allowed but is not standard.
183: The 'e' field specifies the minimum number of digits or spaces in the
184: exponent field on output.
185: If the value of the exponent is too large, the exponent notation 'e'
186: or 'd' will be dropped from the output to allow one
187: more character position.
188: If this is still not adequate, the 'e' field will be filled with
189: asterisks (*). The default value for 'e' is 2.
190: .PP
191: An additional form of tab control specification has been added.
192: The ANSI standard forms 'TRn', 'TLn', and 'Tn' are supported where n is
193: a positive non-zero number. If 'T' or 'nT' is specified, tabbing will
194: be to the next (or n-th) 8-column tab stop.
195: Thus columns of alphanumerics can be lined up without counting.
196: (See above for a description of the tabbing implementation.)
197: .PP
198: A format control specifier has been added to suppress the newline
199: at the end of the last record of a formatted sequential write. The
200: specifier is a dollar sign ($). It is constrained by the same rules
201: as the colon (:). It is used typically for console prompts.
202: For example:
203:
204: .DS
205: write (*, "('enter value for x: ',$)")
206: read (*,*) x
207: .DE
208: .PP
209: Radices other than 10 can be specified for formatted integer I/O
210: conversion. The specifier is patterned after P, the pre-scale factor for
211: floating point conversion. It remains in effect until another radix is
212: specified or format interpretation is complete. The specifier is defined
213: as [n]R where 2 <= n <= 36. If n is omitted,
214: the default decimal radix is restored.
215: .PP
216: In conjunction with the above, a sign control specifier has been added
217: to cause integer values to be interpreted as unsigned during output
218: conversion. The specifier is SU and remains in effect until another
219: sign control specifier is encountered, or format interpretation is
220: complete. Radix and 'unsigned' specifiers could be used to format
221: a hexadecimal dump, as follows:
222:
223: .DS
224: 2000 format( SU, 16R, 8I10.8)
225: .DE
226:
227: Note: Unsigned integer values greater than (2**30 - 1),
228: i.e. any signed negative value, can not be read by FORTRAN input routines.
229: All internal values will be output correctly.
230: .PP
231: The ANSI standard is ambiguous regarding the definition of a "print" file.
232: Since UNIX has no default "print" file, an additional 'form' specifier
233: is now recognized in the 'open' statement.
234: Specifying "form='print'" implies 'formatted' and enables vertical format
235: control for that logical unit (see above).
236: Vertical format control is interpreted only on sequential formatted writes
237: to a "print" file.
238: .PP
239: The 'inquire' statement will return 'print' in the 'FORM=' string variable
240: for logical units opened as "print" files.
241: It will return -1 for the unit number of an unconnected file.
242: .PP
243: If a logical unit is already open, an 'open' statement including the
244: 'form=' option or the 'blank=' option will do nothing but
245: re-define those options.
246: This instance of the 'open' statement need not include the filename, and
247: must not include a filename if 'unit=' refers to the standard input or outputs.
248: Therefore, to re-define the standard output as a "print" file, use:
249:
250: .DS
251: open (unit=6, form='print')
252: .DE
253: .PP
254: In a 'close' statement, "status='keep'" may be specified for temporary files.
255: This is the default for all other files.
256: Remember to get the file's real name,
257: using 'inquire', if you want to re-open it later.
258: .PP
259: List directed read has been modified to allow input of a string not enclosed
260: in quotes. The string must not start with a digit, and can not contain a
261: separator (, or /) or blank (space or tab). A newline will terminate the
262: string unless escaped with \\. Any string not meeting the above restrictions
263: must be enclosed in quotes (" or ').
264: .PP
265: Internal list-directed I/O has been implemented. During internal list reads,
266: bytes are consummed until the iolist is satisfied, or the 'end-of-file'
267: is reached.
268: During internal list writes, records are filled until the iolist is satisfied.
269: The length of an internal array element should be at least 20 bytes to
270: avoid logical record overflow when writing double precision values.
271: Internal list read was implemented to make command line decoding easier.
272: Internal list write should be avoided.
273: .bp
274: .ce 2
275: Appendix A
276: I/O Library Error Messages
277: .PP
278: The following error messages are generated by the I/O library.
279: The error numbers are returned in the "iostat=" variable if the "err="
280: return is taken. Error numbers < 100 are generated by UNIX. See the
281: UNIX Programmers Manual, introduction to chapter 2.
282: .DS
283: /* 100 */ "error in format"
284: See error message output for the location
285: of the error in the format. Can be caused
286: by more than 10 levels of nested (), or
287: an extremely long format statement.
288:
289: /* 101 */ "illegal unit number"
290: It is illegal to close logical unit 0.
291: Negative unit numbers are not allowed.
292: The upper limit is system dependent.
293:
294: /* 102 */ "formatted io not allowed"
295: The logical unit was opened for
296: unformatted I/O.
297:
298: /* 103 */ "unformatted io not allowed"
299: The logical unit was opened for
300: formatted I/O.
301:
302: /* 104 */ "direct io not allowed"
303: The logical unit was opened for sequential
304: access, or the logical record length was
305: specified as 0.
306:
307: /* 105 */ "sequential io not allowed"
308: The logical unit was opened for direct
309: access I/O.
310:
311: /* 106 */ "can't backspace file"
312: The file associated with the logical unit
313: can't seek. May be a device or a pipe.
314:
315: /* 107 */ "off beginning of record"
316: The format specified a left tab off the
317: beginning of the record.
318:
319: /* 108 */ "can't stat file"
320: The system can't return status information
321: about the file. Perhaps the directory is
322: unreadable.
323:
324: /* 109 */ "no * after repeat count"
325: Repeat counts in list-directed I/O must be
326: followed by an * with no blank spaces.
327:
328: .DE
329: .DS
330: /* 110 */ "off end of record"
331: A formatted write tried to go beyond the
332: logical end-of-record. An unformatted read
333: or write will also cause this.
334:
335: /* 111 */ "truncation failed"
336: The truncation of external sequential files
337: on 'close', 'backspace', or 'rewind' tries
338: to do a copy. It failed. Perhaps the temp
339: file couldn't be created.
340:
341: /* 112 */ "incomprehensible list input"
342: List input has to be just right.
343:
344: /* 113 */ "out of free space"
345: The library dynamically creates buffers for
346: internal use. You ran out of memory for this.
347: Your program is too big!
348:
349: /* 114 */ "unit not connected"
350: The logical unit was not open.
351:
352: /* 115 */ "read unexpected character"
353: Certain format conversions can't tolerate
354: non-numeric data. Logical data must be
355: T or F.
356:
357: /* 116 */ "blank logical input field"
358:
359: /* 117 */ "'new' file exists"
360: You tried to open an existing file with
361: "status='new'".
362:
363: /* 118 */ "can't find 'old' file"
364: You tried to open a non-existent file
365: with "status='old'".
366:
367: /* 119 */ "unknown system error"
368: Shouldn't happen, but .....
369: (Send me a documented example.)
370:
371: /* 120 */ "requires seek ability"
372: Direct access requires seek ability.
373: Sequential unformatted I/O requires seek
374: ability on the file due to the special
375: data structure required. Tabbing left
376: also requires seek ability.
377:
378: /* 121 */ "illegal argument"
379: Certain arguments to 'open', etc. will be
380: checked for legitimacy. Often only non-
381: default forms are looked for.
382: .DE
This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.