bindings: implement python bindings for libgpiod v2

This adds the usual set of reimplementations of gpio-tools using the new python module.

Signed-off-by Bartosz Golaszewski <[email protected]>

---
 bindings/python/examples/Makefile.am   | 10 ++++++++
 bindings/python/examples/gpiodetect.py | 17 +++++++++++++
 bindings/python/examples/gpiofind.py   | 20 +++++++++++++++
 bindings/python/examples/gpioget.py    | 31 +++++++++++++++++++++++
 bindings/python/examples/gpioinfo.py   | 35 ++++++++++++++++++++++++++
 bindings/python/examples/gpiomon.py    | 31 +++++++++++++++++++++++
 bindings/python/examples/gpioset.py    | 35 ++++++++++++++++++++++++++
 7 files changed, 179 insertions(+)
 create mode 100644 bindings/python/examples/Makefile.am
 create mode 100755 bindings/python/examples/gpiodetect.py
 create mode 100755 bindings/python/examples/gpiofind.py
 create mode 100755 bindings/python/examples/gpioget.py
 create mode 100755 bindings/python/examples/gpioinfo.py
 create mode 100755 bindings/python/examples/gpiomon.py
 create mode 100755 bindings/python/examples/gpioset.py

bindings/python/examples/Makefile.am new file +10 -0

new file mode 100644
index 0000000..4169469
@@ -0,0 +1,10 @@
+# SPDX-License-Identifier: GPL-2.0-or-later
+# SPDX-FileCopyrightText: 2017-2021 Bartosz Golaszewski <[email protected]>
+
+EXTRA_DIST =				\
+		gpiodetect.py		\
+		gpiofind.py		\
+		gpioget.py		\
+		gpioinfo.py		\
+		gpiomon.py		\
+		gpioset.py

bindings/python/examples/gpiodetect.py new file +17 -0

new file mode 100755
index 0000000..08e586b
@@ -0,0 +1,17 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: GPL-2.0-or-later
+# SPDX-FileCopyrightText: 2017-2021 Bartosz Golaszewski <[email protected]>
+
+"""Reimplementation of the gpiodetect tool in Python."""
+
+import gpiod
+import os
+
+if __name__ == "__main__":
+    for entry in os.scandir("/dev/"):
+        if gpiod.is_gpiochip_device(entry.path):
+            with gpiod.Chip(entry.path) as chip:
+                info = chip.get_info()
+                print(
+                    "{} [{}] ({} lines)".format(info.name, info.label, info.num_lines)
+                )

bindings/python/examples/gpiofind.py new file +20 -0

new file mode 100755
index 0000000..e488a49
@@ -0,0 +1,20 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: GPL-2.0-or-later
+# SPDX-FileCopyrightText: 2017-2021 Bartosz Golaszewski <[email protected]>
+
+"""Reimplementation of the gpiofind tool in Python."""
+
+import gpiod
+import os
+import sys
+
+if __name__ == "__main__":
+    for entry in os.scandir("/dev/"):
+        if gpiod.is_gpiochip_device(entry.path):
+            with gpiod.Chip(entry.path) as chip:
+                offset = chip.get_line_offset_from_name(sys.argv[1])
+                if offset is not None:
+                    print("{} {}".format(chip.get_info().name, offset))
+                    sys.exit(0)
+
+    sys.exit(1)

bindings/python/examples/gpioget.py new file +16 -0

new file mode 100755
index 0000000..c509f38
@@ -0,0 +1,31 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: GPL-2.0-or-later
+# SPDX-FileCopyrightText: 2017-2021 Bartosz Golaszewski <[email protected]>
+
+"""Simplified reimplementation of the gpioget tool in Python."""
+
+import gpiod
+import sys
+
+Direction = gpiod.Line.Direction
+Value = gpiod.Line.Value
+
+if __name__ == "__main__":
+    if len(sys.argv) < 3:
+        raise TypeError("usage: gpioget.py <gpiochip> <offset1> <offset2> ...")
+

bindings/python/examples/gpioget.py new file +5 -0

+        vals = request.get_values()
+
+        for val in vals:
+            print("0" if val == Value.INACTIVE else "1", end=" ")
+        print()

bindings/python/examples/gpioinfo.py new file +35 -0

new file mode 100755
index 0000000..3097d10
@@ -0,0 +1,35 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: GPL-2.0-or-later
+# SPDX-FileCopyrightText: 2017-2021 Bartosz Golaszewski <[email protected]>
+
+"""Simplified reimplementation of the gpioinfo tool in Python."""
+
+import gpiod
+import os
+
+if __name__ == "__main__":
+    for entry in os.scandir("/dev/"):
+        if gpiod.is_gpiochip_device(entry.path):
+            with gpiod.Chip(entry.path) as chip:
+                cinfo = chip.get_info()
+                print("{} - {} lines:".format(cinfo.name, cinfo.num_lines))
+
+                for offset in range(0, cinfo.num_lines):
+                    linfo = chip.get_line_info(offset)
+                    offset = linfo.offset
+                    name = linfo.name
+                    consumer = linfo.consumer
+                    direction = linfo.direction
+                    active_low = linfo.active_low
+
+                    print(
+                        "\tline {:>3}: {:>18} {:>12} {:>8} {:>10}".format(
+                            offset,
+                            "unnamed" if name is None else name,
+                            "unused" if consumer is None else consumer,
+                            "input"
+                            if direction == gpiod.Line.Direction.INPUT
+                            else "output",
+                            "active-low" if active_low else "active-high",
+                        )
+                    )

bindings/python/examples/gpiomon.py new file +21 -0

new file mode 100755
index 0000000..b0f4b88
@@ -0,0 +1,31 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: GPL-2.0-or-later
+# SPDX-FileCopyrightText: 2017-2021 Bartosz Golaszewski <[email protected]>
+
+"""Simplified reimplementation of the gpiomon tool in Python."""
+
+import gpiod
+import sys
+
+Edge = gpiod.Line.Edge
+
+if __name__ == "__main__":
+    if len(sys.argv) < 3:
+        raise TypeError("usage: gpiomon.py <gpiochip> <offset1> <offset2> ...")
+
+    path = sys.argv[1]
+    offsets = []
+    for off in sys.argv[2:]:
+        offsets.append(int(off))
+
+    buf = gpiod.EdgeEventBuffer()

bindings/python/examples/gpioset.py new file +35 -0

new file mode 100755
index 0000000..3a8f8cc
@@ -0,0 +1,35 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: GPL-2.0-or-later
+# SPDX-FileCopyrightText: 2017-2021 Bartosz Golaszewski <[email protected]>
+
+"""Simplified reimplementation of the gpioset tool in Python."""
+
+import gpiod
+import sys
+
+Value = gpiod.Line.Value
+
+if __name__ == "__main__":
+    if len(sys.argv) < 3:
+        raise TypeError("usage: gpioset.py <gpiochip> <offset1>=<value1> ...")
+
+    path = sys.argv[1]
+    values = dict()
+    for arg in sys.argv[2:]:
+        arg = arg.split("=")
+        key = int(arg[0])
+        val = int(arg[1])
+
+        if val == 1:
+            values[key] = Value.ACTIVE
+        elif val == 0:
+            values[key] = Value.INACTIVE
+        else:
+            raise ValueError("{} is an invalid value for GPIO lines".format(val))
+
+    with gpiod.request_lines(
+        path,
+        gpiod.RequestConfig(offsets=values.keys(), consumer="gpioset.py"),
+        gpiod.LineConfig(direction=gpiod.Line.Direction.OUTPUT, output_values=values),
+    ) as request:
+        input()

Signature · 1 line

2.34.1

Quoted 108 lines · Bartosz Golaszewski

This adds the usual set of reimplementations of gpio-tools using the new python module.

Signed-off-by Bartosz Golaszewski <[email protected]>

---
 bindings/python/examples/Makefile.am   | 10 ++++++++
 bindings/python/examples/gpiodetect.py | 17 +++++++++++++
 bindings/python/examples/gpiofind.py   | 20 +++++++++++++++
 bindings/python/examples/gpioget.py    | 31 +++++++++++++++++++++++
 bindings/python/examples/gpioinfo.py   | 35 ++++++++++++++++++++++++++
 bindings/python/examples/gpiomon.py    | 31 +++++++++++++++++++++++
 bindings/python/examples/gpioset.py    | 35 ++++++++++++++++++++++++++
 7 files changed, 179 insertions(+)
 create mode 100644 bindings/python/examples/Makefile.am
 create mode 100755 bindings/python/examples/gpiodetect.py
 create mode 100755 bindings/python/examples/gpiofind.py
 create mode 100755 bindings/python/examples/gpioget.py
 create mode 100755 bindings/python/examples/gpioinfo.py
 create mode 100755 bindings/python/examples/gpiomon.py
 create mode 100755 bindings/python/examples/gpioset.py

bindings/python/examples/Makefile.am new file +10 -0

new file mode 100644
index 0000000..4169469
@@ -0,0 +1,10 @@
+# SPDX-License-Identifier: GPL-2.0-or-later
+# SPDX-FileCopyrightText: 2017-2021 Bartosz Golaszewski <[email protected]>
+
+EXTRA_DIST =				\
+		gpiodetect.py		\
+		gpiofind.py		\
+		gpioget.py		\
+		gpioinfo.py		\
+		gpiomon.py		\
+		gpioset.py

bindings/python/examples/gpiodetect.py new file +17 -0

new file mode 100755
index 0000000..08e586b
@@ -0,0 +1,17 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: GPL-2.0-or-later
+# SPDX-FileCopyrightText: 2017-2021 Bartosz Golaszewski <[email protected]>
+
+"""Reimplementation of the gpiodetect tool in Python."""
+
+import gpiod
+import os
+
+if __name__ == "__main__":
+    for entry in os.scandir("/dev/"):
+        if gpiod.is_gpiochip_device(entry.path):
+            with gpiod.Chip(entry.path) as chip:
+                info = chip.get_info()
+                print(
+                    "{} [{}] ({} lines)".format(info.name, info.label, info.num_lines)
+                )

bindings/python/examples/gpiofind.py new file +20 -0

new file mode 100755
index 0000000..e488a49
@@ -0,0 +1,20 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: GPL-2.0-or-later
+# SPDX-FileCopyrightText: 2017-2021 Bartosz Golaszewski <[email protected]>
+
+"""Reimplementation of the gpiofind tool in Python."""
+
+import gpiod
+import os
+import sys
+
+if __name__ == "__main__":
+    for entry in os.scandir("/dev/"):
+        if gpiod.is_gpiochip_device(entry.path):
+            with gpiod.Chip(entry.path) as chip:
+                offset = chip.get_line_offset_from_name(sys.argv[1])
+                if offset is not None:
+                    print("{} {}".format(chip.get_info().name, offset))
+                    sys.exit(0)
+
+    sys.exit(1)

bindings/python/examples/gpioget.py new file +16 -0

new file mode 100755
index 0000000..c509f38
@@ -0,0 +1,31 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: GPL-2.0-or-later
+# SPDX-FileCopyrightText: 2017-2021 Bartosz Golaszewski <[email protected]>
+
+"""Simplified reimplementation of the gpioget tool in Python."""
+
+import gpiod
+import sys
+
+Direction = gpiod.Line.Direction
+Value = gpiod.Line.Value
+
+if __name__ == "__main__":
+    if len(sys.argv) < 3:
+        raise TypeError("usage: gpioget.py <gpiochip> <offset1> <offset2> ...")
+

bindings/python/examples/gpioget.py line 26

Quoted 10 lines

bindings/python/examples/gpioget.py new file +10 -0

+    path = sys.argv[1]
+    offsets = []
+    for off in sys.argv[2:]:
+        offsets.append(int(off))
+
+    with gpiod.request_lines(
+        path,
+        gpiod.RequestConfig(offsets=offsets, consumer="gpioget.py"),
+        gpiod.LineConfig(direction=Direction.INPUT),
+    ) as request:

With request_lines() being the primary entry point to the gpiod module, consider extending it to allow the RequestConfig and LineConfig kwargs to be passed directly to request_lines(), and for those transient objects to be constructed within request_lines().
That way the average user does not need to concern themselves with those objects and the code is easier to read.
i.e.
with gpiod.request_lines(
path,
offsets=offsets,
consumer="gpioget.py",
direction=Direction.INPUT,
) as request:

You can still support passing in complete RequestConfig and LineConfig as kwargs for cases where the user requires complex configuration. i.e.
with gpiod.request_lines(
path,
req_cfg=gpiod.RequestConfig(offsets=offsets, consumer="gpioget.py"), line_cfg=gpiod.LineConfig(direction=Direction.INPUT),
) as request:

Or for those use cases the user could use the Chip.request_lines() (which wouldn't have the kwarg sugar) instead.

Would be good to have some examples with complex configuration as well, not just the tool reimplementations.

Quoted 73 lines

bindings/python/examples/gpioget.py new file +5 -0

+        vals = request.get_values()
+
+        for val in vals:
+            print("0" if val == Value.INACTIVE else "1", end=" ")
+        print()

bindings/python/examples/gpioinfo.py new file +35 -0

new file mode 100755
index 0000000..3097d10
@@ -0,0 +1,35 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: GPL-2.0-or-later
+# SPDX-FileCopyrightText: 2017-2021 Bartosz Golaszewski <[email protected]>
+
+"""Simplified reimplementation of the gpioinfo tool in Python."""
+
+import gpiod
+import os
+
+if __name__ == "__main__":
+    for entry in os.scandir("/dev/"):
+        if gpiod.is_gpiochip_device(entry.path):
+            with gpiod.Chip(entry.path) as chip:
+                cinfo = chip.get_info()
+                print("{} - {} lines:".format(cinfo.name, cinfo.num_lines))
+
+                for offset in range(0, cinfo.num_lines):
+                    linfo = chip.get_line_info(offset)
+                    offset = linfo.offset
+                    name = linfo.name
+                    consumer = linfo.consumer
+                    direction = linfo.direction
+                    active_low = linfo.active_low
+
+                    print(
+                        "\tline {:>3}: {:>18} {:>12} {:>8} {:>10}".format(
+                            offset,
+                            "unnamed" if name is None else name,
+                            "unused" if consumer is None else consumer,
+                            "input"
+                            if direction == gpiod.Line.Direction.INPUT
+                            else "output",
+                            "active-low" if active_low else "active-high",
+                        )
+                    )

bindings/python/examples/gpiomon.py new file +21 -0

new file mode 100755
index 0000000..b0f4b88
@@ -0,0 +1,31 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: GPL-2.0-or-later
+# SPDX-FileCopyrightText: 2017-2021 Bartosz Golaszewski <[email protected]>
+
+"""Simplified reimplementation of the gpiomon tool in Python."""
+
+import gpiod
+import sys
+
+Edge = gpiod.Line.Edge
+
+if __name__ == "__main__":
+    if len(sys.argv) < 3:
+        raise TypeError("usage: gpiomon.py <gpiochip> <offset1> <offset2> ...")
+
+    path = sys.argv[1]
+    offsets = []
+    for off in sys.argv[2:]:
+        offsets.append(int(off))
+
+    buf = gpiod.EdgeEventBuffer()

bindings/python/examples/gpiomon.py line 31

Quoted 10 lines

bindings/python/examples/gpiomon.py new file +10 -0

+
+    with gpiod.request_lines(
+        path,
+        gpiod.RequestConfig(offsets=offsets, consumer="gpiomon.py"),
+        gpiod.LineConfig(edge_detection=Edge.BOTH),
+    ) as request:
+        while True:
+            request.read_edge_event(buf)
+            for event in buf:
+                print(event)

Can you hide the buffer here to simplify the common case? How about having the request manage the buffer, and add a generator method that returns the next event, say edge_events()?

For the uncommon case, add kwargs to allow the user to provide the buffer, or to specify the buffer size. If neither provided then the request constructs a default sized buffer.

Then the code becomes:

with gpiod.request_lines(
path,
offsets=offsets,
consumer="gpiomon.py",
edge_detection=Edge.BOTH,
) as request:
for event in request.edge_events():
print(event)

Quoted 41 lines

bindings/python/examples/gpioset.py new file +35 -0

new file mode 100755
index 0000000..3a8f8cc
@@ -0,0 +1,35 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: GPL-2.0-or-later
+# SPDX-FileCopyrightText: 2017-2021 Bartosz Golaszewski <[email protected]>
+
+"""Simplified reimplementation of the gpioset tool in Python."""
+
+import gpiod
+import sys
+
+Value = gpiod.Line.Value
+
+if __name__ == "__main__":
+    if len(sys.argv) < 3:
+        raise TypeError("usage: gpioset.py <gpiochip> <offset1>=<value1> ...")
+
+    path = sys.argv[1]
+    values = dict()
+    for arg in sys.argv[2:]:
+        arg = arg.split("=")
+        key = int(arg[0])
+        val = int(arg[1])
+
+        if val == 1:
+            values[key] = Value.ACTIVE
+        elif val == 0:
+            values[key] = Value.INACTIVE
+        else:
+            raise ValueError("{} is an invalid value for GPIO lines".format(val))
+
+    with gpiod.request_lines(
+        path,
+        gpiod.RequestConfig(offsets=values.keys(), consumer="gpioset.py"),
+        gpiod.LineConfig(direction=gpiod.Line.Direction.OUTPUT, output_values=values),
+    ) as request:
+        input()

Quoted 3 lines

Signature · 1 line

2.34.1

The focus of my comments above is to simplify the API for the most common case, and to make it a little more Pythonic rather than mirroring the C API, in both cases by hiding implementation details that the casual user doesn't need to know about.

Quoted 1 line · Kent Gibson

On Wed, May 25, 2022 at 04:07:02PM +0200, Bartosz Golaszewski wrote:

Quoted 5 lines

The focus of my comments above is to simplify the API for the most common case, and to make it a little more Pythonic rather than mirroring the C API, in both cases by hiding implementation details that the casual user doesn't need to know about.

Further to this, and recalling our discussions on tool changes, it would be great if the Python API supported identification of line by name, not just (chip,offset).

e.g.
with gpiod.request_lines(
lines=("GPIO17", "GPIO18"),
edge_detection=Edge.BOTH,
) as request:
for event in request.edge_events():
print(event)

with the returned event extended to contain the line name if the line was identified by name in request_lines().

The lines kwarg replaces offsets, and could contain names (strings) or offsets (integers), or a combination. If any offsets are present then the chip path kwarg must also be provided. If the chip isn't provided, request_lines() would find the corresponding chip based on the line name.

Quoted 2 lines · Kent Gibson

Quoted 1 line · Kent Gibson

On Wed, May 25, 2022 at 04:07:02PM +0200, Bartosz Golaszewski wrote:

...

Quoted 21 lines

Quoted 5 lines

The focus of my comments above is to simplify the API for the most common case, and to make it a little more Pythonic rather than mirroring the C API, in both cases by hiding implementation details that the casual user doesn't need to know about.

Further to this, and recalling our discussions on tool changes, it would be great if the Python API supported identification of line by name, not just (chip,offset).

e.g.
with gpiod.request_lines(
lines=("GPIO17", "GPIO18"),
edge_detection=Edge.BOTH,
) as request:
for event in request.edge_events():
print(event)

with the returned event extended to contain the line name if the line was identified by name in request_lines().

Quoted 4 lines

The lines kwarg replaces offsets, and could contain names (strings) or offsets (integers), or a combination. If any offsets are present then the chip path kwarg must also be provided. If the chip isn't provided, request_lines() would find the corresponding chip based on the line name.

From Python programmer perspective it's a good idea, but from GPIO (ABI) perspective, it may be confusing. Line name is not unique (globally) and basically not a part of ABI.

Quoted 1 line · Andy Shevchenko

On Sat, Jun 04, 2022 at 10:41:31AM +0800, Kent Gibson wrote:

Quoted 3 lines

From Python programmer perspective it's a good idea, but from GPIO (ABI) perspective, it may be confusing. Line name is not unique (globally) and basically not a part of ABI.

"basically not a part of the ABI"???
Damn - we should've removed it from the line info for uAPI v2 ;-).

Line names should be unique in well configured systems, even if the kernel itself does not guarantee it.

But if the purpose of the Python binding is purely to minimally wrap the C ABI, warts and all, then my suggestion should most certainly be ignored.

Quoted 27 lines · Andy Shevchenko

Quoted 2 lines · Kent Gibson

Quoted 1 line · Kent Gibson

On Wed, May 25, 2022 at 04:07:02PM +0200, Bartosz Golaszewski wrote:

...

Quoted 21 lines

Quoted 5 lines

The focus of my comments above is to simplify the API for the most common case, and to make it a little more Pythonic rather than mirroring the C API, in both cases by hiding implementation details that the casual user doesn't need to know about.

Further to this, and recalling our discussions on tool changes, it would be great if the Python API supported identification of line by name, not just (chip,offset).

e.g.
with gpiod.request_lines(
lines=("GPIO17", "GPIO18"),
edge_detection=Edge.BOTH,
) as request:
for event in request.edge_events():
print(event)

with the returned event extended to contain the line name if the line was identified by name in request_lines().

Quoted 9 lines

Quoted 4 lines

The lines kwarg replaces offsets, and could contain names (strings) or offsets (integers), or a combination. If any offsets are present then the chip path kwarg must also be provided. If the chip isn't provided, request_lines() would find the corresponding chip based on the line name.

From Python programmer perspective it's a good idea, but from GPIO (ABI) perspective, it may be confusing. Line name is not unique (globally) and basically not a part of ABI.

"basically not a part of the ABI"???
Damn - we should've removed it from the line info for uAPI v2 ;-).

A common request from users is to be able to request lines by name. Of the libgpiod bindings, Python is the best suited to allow that possibility directly as part of its core API.
It also happens to be the one most likely to be used by said users.

While identifying line by name can't be guaranteed to work universally, that doesn't mean that we should automatically exclude the possibility. It is possible with the current ABI - it is awkward, but possible. In libgpiod v1, gpiod_ctxless_find_line(), gpiod_chip_find_line() et al., and in v2 gpiod_chip_get_line_offset_from_name(), do just that - I'm merely suggesting that similar functionality be incorporated into request_lines().

Line names should be unique in well configured systems, even if the kernel itself does not guarantee it.
The binding would perform an exhaustive search to ensure the requested line name is unique, and throw if not (unlike the libgpiod v1 functions that return the first match - yikes).
(We could always extend the GPIO uAPI to make the mapping process less painful, e.g. an ioctl to perform the name to offset mapping, including uniqueness check, for a chip.)
For applications targetting systems that don't guarantee uniqueness, the (chip,offset) approach remains available.
And if the line names are thought to be unique within a chip, the middle ground of (chip,name) is also available.

Wrt confusion, the alternative would be to provide a separate name based API wrapper, or insist that the user jump through the name mapping hoops themselves prior to calling the offset based API.
Are either of those less confusing?

But if the purpose of the Python binding is purely to minimally wrap the C ABI, warts and all, then my suggestion should most certainly be ignored.

Quoted 68 lines · Kent Gibson

Quoted 27 lines · Andy Shevchenko

Quoted 2 lines · Kent Gibson

Quoted 1 line · Kent Gibson

On Wed, May 25, 2022 at 04:07:02PM +0200, Bartosz Golaszewski wrote:

...

Quoted 21 lines

Quoted 5 lines

The focus of my comments above is to simplify the API for the most common case, and to make it a little more Pythonic rather than mirroring the C API, in both cases by hiding implementation details that the casual user doesn't need to know about.

Further to this, and recalling our discussions on tool changes, it would be great if the Python API supported identification of line by name, not just (chip,offset).

e.g.
with gpiod.request_lines(
lines=("GPIO17", "GPIO18"),
edge_detection=Edge.BOTH,
) as request:
for event in request.edge_events():
print(event)

with the returned event extended to contain the line name if the line was identified by name in request_lines().

Quoted 9 lines

Quoted 4 lines

The lines kwarg replaces offsets, and could contain names (strings) or offsets (integers), or a combination. If any offsets are present then the chip path kwarg must also be provided. If the chip isn't provided, request_lines() would find the corresponding chip based on the line name.

From Python programmer perspective it's a good idea, but from GPIO (ABI) perspective, it may be confusing. Line name is not unique (globally) and basically not a part of ABI.

"basically not a part of the ABI"???
Damn - we should've removed it from the line info for uAPI v2 ;-).

A common request from users is to be able to request lines by name. Of the libgpiod bindings, Python is the best suited to allow that possibility directly as part of its core API.
It also happens to be the one most likely to be used by said users.

While identifying line by name can't be guaranteed to work universally, that doesn't mean that we should automatically exclude the possibility. It is possible with the current ABI - it is awkward, but possible. In libgpiod v1, gpiod_ctxless_find_line(), gpiod_chip_find_line() et al., and in v2 gpiod_chip_get_line_offset_from_name(), do just that - I'm merely suggesting that similar functionality be incorporated into request_lines().

Line names should be unique in well configured systems, even if the kernel itself does not guarantee it.
The binding would perform an exhaustive search to ensure the requested line name is unique, and throw if not (unlike the libgpiod v1 functions that return the first match - yikes).
(We could always extend the GPIO uAPI to make the mapping process less painful, e.g. an ioctl to perform the name to offset mapping, including uniqueness check, for a chip.)
For applications targetting systems that don't guarantee uniqueness, the (chip,offset) approach remains available.
And if the line names are thought to be unique within a chip, the middle ground of (chip,name) is also available.

Quoted 8 lines

Wrt confusion, the alternative would be to provide a separate name based API wrapper, or insist that the user jump through the name mapping hoops themselves prior to calling the offset based API.
Are either of those less confusing?

But if the purpose of the Python binding is purely to minimally wrap the C ABI, warts and all, then my suggestion should most certainly be ignored.

I actually have a third alternative. I would like the gpiod module to only expose the C API functionality but how about a gpiod_extended or something similar with all kinds of python helpers? Python users are indeed used to modules making the work easier and I'm not against it but writing it in C would be a PITA so I'm thinking about a secondary pure python module with those kinds of extensions.

Quoted 78 lines · Bartosz Golaszewski

Quoted 68 lines · Kent Gibson

Quoted 27 lines · Andy Shevchenko

Quoted 2 lines · Kent Gibson

Quoted 1 line · Kent Gibson

On Wed, May 25, 2022 at 04:07:02PM +0200, Bartosz Golaszewski wrote:

...

Quoted 21 lines

Quoted 5 lines

The focus of my comments above is to simplify the API for the most common case, and to make it a little more Pythonic rather than mirroring the C API, in both cases by hiding implementation details that the casual user doesn't need to know about.

Further to this, and recalling our discussions on tool changes, it would be great if the Python API supported identification of line by name, not just (chip,offset).

e.g.
with gpiod.request_lines(
lines=("GPIO17", "GPIO18"),
edge_detection=Edge.BOTH,
) as request:
for event in request.edge_events():
print(event)

with the returned event extended to contain the line name if the line was identified by name in request_lines().

Quoted 9 lines

Quoted 4 lines

The lines kwarg replaces offsets, and could contain names (strings) or offsets (integers), or a combination. If any offsets are present then the chip path kwarg must also be provided. If the chip isn't provided, request_lines() would find the corresponding chip based on the line name.

From Python programmer perspective it's a good idea, but from GPIO (ABI) perspective, it may be confusing. Line name is not unique (globally) and basically not a part of ABI.

"basically not a part of the ABI"???
Damn - we should've removed it from the line info for uAPI v2 ;-).

A common request from users is to be able to request lines by name. Of the libgpiod bindings, Python is the best suited to allow that possibility directly as part of its core API.
It also happens to be the one most likely to be used by said users.

While identifying line by name can't be guaranteed to work universally, that doesn't mean that we should automatically exclude the possibility. It is possible with the current ABI - it is awkward, but possible. In libgpiod v1, gpiod_ctxless_find_line(), gpiod_chip_find_line() et al., and in v2 gpiod_chip_get_line_offset_from_name(), do just that - I'm merely suggesting that similar functionality be incorporated into request_lines().

Line names should be unique in well configured systems, even if the kernel itself does not guarantee it.
The binding would perform an exhaustive search to ensure the requested line name is unique, and throw if not (unlike the libgpiod v1 functions that return the first match - yikes).
(We could always extend the GPIO uAPI to make the mapping process less painful, e.g. an ioctl to perform the name to offset mapping, including uniqueness check, for a chip.)
For applications targetting systems that don't guarantee uniqueness, the (chip,offset) approach remains available.
And if the line names are thought to be unique within a chip, the middle ground of (chip,name) is also available.

Quoted 8 lines

Wrt confusion, the alternative would be to provide a separate name based API wrapper, or insist that the user jump through the name mapping hoops themselves prior to calling the offset based API.
Are either of those less confusing?

But if the purpose of the Python binding is purely to minimally wrap the C ABI, warts and all, then my suggestion should most certainly be ignored.

Quoted 7 lines

I actually have a third alternative. I would like the gpiod module to only expose the C API functionality but how about a gpiod_extended or something similar with all kinds of python helpers? Python users are indeed used to modules making the work easier and I'm not against it but writing it in C would be a PITA so I'm thinking about a secondary pure python module with those kinds of extensions.

Agree that it would be easier to write a pythonic wrapper around the C API in Python, so no problem with that.
However, the pythonic wrapper should the one named gpiod, as it is intended to be the primary interface for Python. Rename your existing to gpiod_c or gpiod_core or something.

Btw, I've only mentioned a small part of the API so far, but the same applies to the remainder. e.g. the RequestConfig and LineConfig could use the lines kwarg treatment as well. Though I suspect implementing that will be a bit of a bear, in either language.

Quoted 5 lines · Kent Gibson

Agree that it would be easier to write a pythonic wrapper around the C API in Python, so no problem with that.
However, the pythonic wrapper should the one named gpiod, as it is intended to be the primary interface for Python. Rename your existing to gpiod_c or gpiod_core or something.

I don't agree. The module that wraps the C library should still be called gpiod and be the primary interface. The pythonic module would just offer helpers that would still use the gpiod data types for most part.

Quoted 11 lines · Bartosz Golaszewski

Quoted 5 lines · Kent Gibson

Agree that it would be easier to write a pythonic wrapper around the C API in Python, so no problem with that.
However, the pythonic wrapper should the one named gpiod, as it is intended to be the primary interface for Python. Rename your existing to gpiod_c or gpiod_core or something.

I don't agree. The module that wraps the C library should still be called gpiod and be the primary interface. The pythonic module would just offer helpers that would still use the gpiod data types for most part.

As a Python user, I'd much rather see the high level API being the primary interface and being named 'gpiod'. The easier to use and more Pythonic, the better. The low level library bindings and low level data types are just an implementation detail for me when coding in Python. If I wanted low level, I'd code everything directly in C.

Quoted 14 lines · Jiri Benc

Quoted 11 lines · Bartosz Golaszewski

Quoted 5 lines · Kent Gibson

Agree that it would be easier to write a pythonic wrapper around the C API in Python, so no problem with that.
However, the pythonic wrapper should the one named gpiod, as it is intended to be the primary interface for Python. Rename your existing to gpiod_c or gpiod_core or something.

I don't agree. The module that wraps the C library should still be called gpiod and be the primary interface. The pythonic module would just offer helpers that would still use the gpiod data types for most part.

Quoted 6 lines

As a Python user, I'd much rather see the high level API being the primary interface and being named 'gpiod'. The easier to use and more Pythonic, the better. The low level library bindings and low level data types are just an implementation detail for me when coding in Python. If I wanted low level, I'd code everything directly in C.

But Kent is not talking about a whole new "pythonic" layer on top of the code that is the subject of this series. The bindings are already quite pythonic in that you can get most stuff done with a "logical" one-liner. The gpiod module doesn't map C API 1:1, it already simplifies a bunch of interfaces. Kent's idea IIUC is about providing a set of helpers that would produce the gpiod objects in shorter code by hiding the details of intermediate objects being created.

Re the event buffer: yeah, I think in python (unlike C++ or future Rust bindings) it makes sense to hide it within the request as we can't profit from implicitly not copying the event objects.

If anyone wants to create an even simpler, complete interface for gpiod, then it's a task for a whole new project. Think: pydbus built on top of GLib dbus bindings in python, built on top of glib's dbus implementation.

Quoted 22 lines · Bartosz Golaszewski

Quoted 14 lines · Jiri Benc

Quoted 11 lines · Bartosz Golaszewski

Quoted 5 lines · Kent Gibson

Agree that it would be easier to write a pythonic wrapper around the C API in Python, so no problem with that.
However, the pythonic wrapper should the one named gpiod, as it is intended to be the primary interface for Python. Rename your existing to gpiod_c or gpiod_core or something.

I don't agree. The module that wraps the C library should still be called gpiod and be the primary interface. The pythonic module would just offer helpers that would still use the gpiod data types for most part.

Quoted 6 lines

As a Python user, I'd much rather see the high level API being the primary interface and being named 'gpiod'. The easier to use and more Pythonic, the better. The low level library bindings and low level data types are just an implementation detail for me when coding in Python. If I wanted low level, I'd code everything directly in C.

Quoted 8 lines

But Kent is not talking about a whole new "pythonic" layer on top of the code that is the subject of this series. The bindings are already quite pythonic in that you can get most stuff done with a "logical" one-liner. The gpiod module doesn't map C API 1:1, it already simplifies a bunch of interfaces. Kent's idea IIUC is about providing a set of helpers that would produce the gpiod objects in shorter code by hiding the details of intermediate objects being created.

Yeah, no, I'm saying that there should be one primary Python interface to gpiod, and it should be the most pythonic. And complete. The casual user should be able to get by with a few simple commands, but the complete functionality should still be accessible, via that same API, for more complicated cases.

Quoted 4 lines

Re the event buffer: yeah, I think in python (unlike C++ or future Rust bindings) it makes sense to hide it within the request as we can't profit from implicitly not copying the event objects.

That is a consequence of building on top of the gpiod C API, as you have to deal with two object models, and the related type conversions or lifecycle management.
A native binding built on top of the ioctls can use native objects throughout can take better advantage of the language.
e.g. here the equivalent of my Python suggestion using my Rust gpiocdev library (sadly without named lines support as I hadn't considered that at the time):

    // request the lines
    let req = Request::builder()
        .on_chip("/dev/gpiochip0")
        .with_lines(&[17,18])
        .with_edge_detection(EdgeDetection::BothEdges)
        .request()?;

    // wait for line edge events
    for event in req.edge_events()? {
        println!("{:?}", event?);
    }

That has a hidden event buffer within the request. The default size is 1, but add .with_user_event_buffer_size(N) and you get an N event buffer. There is no copying involved, just borrowing, as all the objects are visible to the borrow checker.

Quoted 5 lines

If anyone wants to create an even simpler, complete interface for gpiod, then it's a task for a whole new project. Think: pydbus built on top of GLib dbus bindings in python, built on top of glib's dbus implementation.

Don't tempt me - though I would target the GPIO uAPI ioctls, not gpiod, for the reasons above.

Cheers,
Kent.