/usr/share/doc/python-apsw/html/_sources/cursor.txt is in python-apsw-doc 3.7.6.3-r1-1build1.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 | .. Automatically generated by code2rst.py
code2rst.py src/cursor.c doc/cursor.rst
Edit src/cursor.c not this file!
.. currentmodule:: apsw
.. _cursors:
Cursors (executing SQL)
***********************
A cursor encapsulates a SQL query and returning results. To make a
new cursor you should call :meth:`~Connection.cursor` on your
database::
db=apsw.Connection("databasefilename")
cursor=db.cursor()
A cursor executes SQL::
cursor.execute("create table example(title, isbn)")
You can also read data back. The row is returned as a tuple of the
column values::
for row in cursor.execute("select * from example"):
print row
There are two ways of supplying data to a query. The **really bad** way is to compose a string::
sql="insert into example values('%s', %d)" % ("string", 8390823904)
cursor.execute(sql)
If there were any single quotes in string then you would have invalid
syntax. Additionally this is how `SQL injection attacks
<http://en.wikipedia.org/wiki/SQL_injection>`_ happen. Instead you should use bindings::
sql="insert into example values(?, ?)"
cursor.execute(sql, ("string", 8390823904))
# You can also use dictionaries
sql="insert into example values(:title, :isbn)"
cursor.execute(sql, {"title": "string", "isbn": 8390823904})
# You can use local variables as the dictionary
title="..."
isbn="...."
cursor.execute(sql, locals())
Cursors are cheap. Use as many as you need. It is safe to use them
across threads, such as calling :meth:`~Cursor.execute` in one thread,
passing the cursor to another thread that then calls
:meth:`Cursor.next`. The only thing you can't do is call methods at
exactly the same time on the same cursor in two different threads - eg
trying to call :meth:`~Cursor.execute` in both at the same time, or
:meth:`~Cursor.execute` in one and :meth:`Cursor.next` in another.
(If you do attempt this, it will be detected and
:exc:`ThreadingViolationError` will be raised.)
Behind the scenes a :class:`Cursor` maps to a `SQLite statement
<http://www.sqlite.org/c3ref/stmt.html>`_. APSW maintains a
:ref:`cache <statementcache>` so that the mapping is very fast, and the
SQLite objects are reused when possible.
A unique feature of APSW is that your query can be multiple semi-colon
separated statements. For example::
cursor.execute("select ... ; insert into ... ; update ... ; select ...")
.. note::
SQLite fetches data as it is needed. If table *example* had 10
million rows it would only get the next row as requested (the for
loop effectively calls :meth:`~Cursor.next` to get each row). This
code would not work as expected::
for row in cursor.execute("select * from example"):
cursor.execute("insert .....")
The nested :meth:`~Cursor.execute` would start a new query
abandoning any remaining results from the ``SELECT`` cursor. There are two
ways to work around this. Use a different cursor::
for row in cursor1.execute("select * from example"):
cursor2.execute("insert ...")
You can also get all the rows immediately by filling in a list::
rows=list( cursor.execute("select * from example") )
for row in rows:
cursor.execute("insert ...")
This last approach is recommended since you don't have to worry
about the database changing while doing the ``select``. You should
also understand transactions and where to put the transaction
boundaries.
.. note::
Cursors on the same :ref:`Connection <connections>` are not isolated
from each other. Anything done on one cursor is immediately visible
to all other Cursors on the same connection. This still applies if
you start transactions. Connections are isolated from each other
with cursors on other connections not seeing changes until they are
committed.
.. seealso::
* `SQLite transactions <http://www.sqlite.org/lang_transaction.html>`_
* `Atomic commit <http://www.sqlite.org/atomiccommit.html>`_
* `Example of changing the database while running a query problem <http://www.mail-archive.com/sqlite-users@sqlite.org/msg42660.html>`_
* :ref:`Benchmarking`
Cursor class
============
.. class:: Cursor
You obtain cursors by calling :meth:`Connection.cursor`.
.. method:: Cursor.close(force=False)
It is very unlikely you will need to call this method. It exists
because older versions of SQLite required all Connection/Cursor
activity to be confined to the same thread. That is no longer the
case. Cursors are automatically garbage collected and when there
are none left will allow the connection to be garbage collected if
it has no other references.
A cursor is open if there are remaining statements to execute (if
your query included multiple statements), or if you called
:meth:`~Cursor.executemany` and not all of the *sequenceofbindings*
have been used yet.
:param force: If False then you will get exceptions if there is
remaining work to do be in the Cursor such as more statements to
execute, more data from the executemany binding sequence etc. If
force is True then all remaining work and state information will be
silently discarded.
.. index:: sqlite3_prepare_v2, sqlite3_step, sqlite3_bind_int64, sqlite3_bind_null, sqlite3_bind_text, sqlite3_bind_double, sqlite3_bind_blob, sqlite3_bind_zeroblob
.. method:: Cursor.execute(statements[, bindings]) -> iterator
Executes the statements using the supplied bindings. Execution
returns when the first row is available or all statements have
completed.
:param statements: One or more SQL statements such as ``select *
from books`` or ``begin; insert into books ...; select
last_insert_rowid(); end``.
:param bindings: If supplied should either be a sequence or a dictionary. Each item must be one of the :ref:`supported types <types>`
If you use numbered bindings in the query then supply a sequence.
Any sequence will work including lists and iterators. For
example::
cursor.execute("insert into books values(?,?)", ("title", "number"))
.. note::
A common gotcha is wanting to insert a single string but not
putting it in a tuple::
cursor.execute("insert into books values(?)", "a title")
The string is a sequence of 8 characters and so it will look
like you are supplying 8 bindings when only one is needed. Use
a one item tuple with a trailing comma like this::
cursor.execute("insert into books values(?)", ("a title",) )
If you used names in the statement then supply a dictionary as the
binding. It is ok to be missing entries from the dictionary -
None/null will be used. For example::
cursor.execute("insert into books values(:title, :isbn, :rating)",
{"title": "book title", "isbn": 908908908})
The return is the cursor object itself which is also an iterator. This allows you to write::
for row in cursor.execute("select * from books"):
print row
:raises TypeError: The bindings supplied were neither a dict nor a sequence
:raises BindingsError: You supplied too many or too few bindings for the statements
:raises IncompleteExecutionError: There are remaining unexecuted queries from your last execute
.. seealso::
* :ref:`executionmodel`
* :ref:`Example <example-cursor>`
Calls:
* :sqliteapi:`sqlite3_prepare_v2 <prepare>`
* :sqliteapi:`sqlite3_step <step>`
* :sqliteapi:`sqlite3_bind_int64 <bind_blob>`
* :sqliteapi:`sqlite3_bind_null <bind_blob>`
* :sqliteapi:`sqlite3_bind_text <bind_blob>`
* :sqliteapi:`sqlite3_bind_double <bind_blob>`
* :sqliteapi:`sqlite3_bind_blob <bind_blob>`
* :sqliteapi:`sqlite3_bind_zeroblob <bind_blob>`
.. method:: Cursor.executemany(statements, sequenceofbindings) -> iterator
This method is for when you want to execute the same statements over
a sequence of bindings. Conceptually it does this::
for binding in sequenceofbindings:
cursor.execute(statements, binding)
Example::
rows=( (1, 7),
(2, 23),
(4, 92),
(12, 12) )
cursor.executemany("insert into nums values(?,?)", rows)
The return is the cursor itself which acts as an iterator. Your
statements can return data. See :meth:`~Cursor.execute` for more
information.
.. method:: Cursor.fetchall() -> list
Returns all remaining result rows as a list. This method is defined
in DBAPI. It is a longer way of doing ``list(cursor)``.
.. method:: Cursor.getconnection() -> Connection
Returns the :class:`Connection` this cursor belongs to. An example usage is to get another cursor::
def func(cursor):
# I don't want to alter existing cursor, so make a new one
mycursor=cursor.getconnection().cursor()
mycursor.execute("....")
.. index:: sqlite3_column_name, sqlite3_column_decltype
.. method:: Cursor.getdescription() -> list
Returns a list describing each column in the result row. The
return is identical for every row of the results. You can only
call this method once you have started executing a statement and
before you have finished::
# This will error
cursor.getdescription()
for row in cursor.execute("select ....."):
# this works
print cursor.getdescription()
print row
The information about each column is a tuple of ``(column_name,
declared_column_type)``. The type is what was declared in the
``CREATE TABLE`` statement - the value returned in the row will be
whatever type you put in for that row and column. (This is known
as `manifest typing <http://www.sqlite.org/different.html#typing>`_
which is also the way that Python works. The variable ``a`` could
contain an integer, and then you could put a string in it. Other
static languages such as C or other SQL databases only let you put
one type in - eg ``a`` could only contain an integer or a string,
but never both.)
Example::
cursor.execute("create table books(title string, isbn number, wibbly wobbly zebra)")
cursor.execute("insert into books values(?,?,?)", (97, "fjfjfj", 3.7))
cursor.execute("insert into books values(?,?,?)", ("fjfjfj", 3.7, 97))
for row in cursor.execute("select * from books"):
print cursor.getdescription()
print row
Output::
# row 0 - description
(('title', 'string'), ('isbn', 'number'), ('wibbly', 'wobbly zebra'))
# row 0 - values
(97, 'fjfjfj', 3.7)
# row 1 - description
(('title', 'string'), ('isbn', 'number'), ('wibbly', 'wobbly zebra'))
# row 1 - values
('fjfjfj', 3.7, 97)
Calls:
* :sqliteapi:`sqlite3_column_name <column_name>`
* :sqliteapi:`sqlite3_column_decltype <column_decltype>`
.. method:: Cursor.getexectrace() -> callable or None
Returns the currently installed (via :meth:`~Cursor.setexectrace`)
execution tracer.
.. seealso::
* :ref:`tracing`
.. method:: Cursor.getrowtrace() -> callable or None
Returns the currently installed (via :meth:`~Cursor.setrowtrace`)
row tracer.
.. seealso::
* :ref:`tracing`
.. method:: Cursor.next() -> row
Returns the next row of data or raises StopIteration if there are no
more rows. Python calls this method behind the scenes when using
the cursor as an iterator. It is unlikely you will want to manually
call it.
.. method:: Cursor.setexectrace(callable)
*callable* is called with the cursor, statement and bindings for
each :meth:`~Cursor.execute` or :meth:`~Cursor.executemany` on this
cursor.
If *callable* is :const:`None` then any existing execution tracer is
removed.
.. seealso::
* :ref:`tracing`
* :ref:`executiontracer`
* :meth:`Connection.setexectrace`
.. method:: Cursor.setrowtrace(callable)
*callable* is called with cursor and row being returned. You can
change the data that is returned or cause the row to be skipped
altogether.
If *callable* is :const:`None` then any existing row tracer is
removed.
.. seealso::
* :ref:`tracing`
* :ref:`rowtracer`
* :meth:`Connection.setexectrace`
|